| Age | Commit message (Collapse) | Author |
|---|
|
When draining an NBD export, nbd_drained_begin() first sets
client->quiescing so that nbd_client_receive_next_request() won't start
any new request coroutines. Then nbd_drained_poll() tries to makes sure
that we wait for any existing request coroutines by checking that
client->nb_requests has become 0.
However, there is a small window between creating a new request
coroutine and increasing client->nb_requests. If a coroutine is in this
state, it won't be waited for and drain returns too early.
In the context of switching to a different AioContext, this means that
blk_aio_attached() will see client->recv_coroutine != NULL and fail its
assertion.
Fix this by increasing client->nb_requests immediately when starting the
coroutine. Doing this after the checks if we should create a new
coroutine is okay because client->lock is held.
Cc: qemu-stable@nongnu.org
Fixes: fd6afc501a01 ("nbd/server: Use drained block ops to quiesce the server")
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20240314165825.40261-2-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
NBDClient has a number of fields that are accessed by both the export
AioContext and the main loop thread. When the AioContext lock is removed
these fields will need another form of protection.
Add NBDClient->lock and protect fields that are accessed by both
threads. Also add assertions where possible and otherwise add doc
comments stating assumptions about which thread and lock holding.
Note this patch moves the client->recv_coroutine assertion from
nbd_co_receive_request() to nbd_trip() where client->lock is held.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20231221192452.1785567-7-stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
The NBD clients list is currently accessed from both the export
AioContext and the main loop thread. When the AioContext lock is removed
there will be nothing protecting the clients list.
Adding a lock around the clients list is tricky because NBDClient
structs are refcounted and may be freed from the export AioContext or
the main loop thread. nbd_export_request_shutdown() -> client_close() ->
nbd_client_put() is also tricky because the list lock would be held
while indirectly dropping references to NDBClients.
A simpler approach is to only allow nbd_client_put() and client_close()
calls from the main loop thread. Then the NBD clients list is only
accessed from the main loop thread and no fancy locking is needed.
nbd_trip() just needs to reschedule itself in the main loop AioContext
before calling nbd_client_put() and client_close(). This costs more CPU
cycles per NBD request so add nbd_client_put_nonzero() to optimize the
common case where more references to NBDClient remain.
Note that nbd_client_get() can still be called from either thread, so
make NBDClient->refcount atomic.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20231221192452.1785567-6-stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
nbd_trip() processes a single NBD request from start to finish and holds
an NBDClient reference throughout. NBDRequest does not outlive the scope
of nbd_trip(). Therefore it is unnecessary to ref/unref NBDClient for
each NBDRequest.
Removing these nbd_client_get()/nbd_client_put() calls will make
thread-safety easier in the commits that follow.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <20231221192452.1785567-5-stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
This adds GRAPH_RDLOCK annotations to declare that callers of
bdrv_filter_or_cow_bs() need to hold a reader lock for the graph because
it calls bdrv_filter_or_cow_child(), which accesses bs->file/backing.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20231027155333.420094-7-kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
Allow a client to request a subset of negotiated meta contexts. For
example, a client may ask to use a single connection to learn about
both block status and dirty bitmaps, but where the dirty bitmap
queries only need to be performed on a subset of the disk; forcing the
server to compute that information on block status queries in the rest
of the disk is wasted effort (both at the server, and on the amount of
traffic sent over the wire to be parsed and ignored by the client).
Qemu as an NBD client never requests to use more than one meta
context, so it has no need to use block status payloads. Testing this
instead requires support from libnbd, which CAN access multiple meta
contexts in parallel from a single NBD connection; an interop test
submitted to the libnbd project at the same time as this patch
demonstrates the feature working, as well as testing some corner cases
(for example, when the payload length is longer than the export
length), although other corner cases (like passing the same id
duplicated) requires a protocol fuzzer because libnbd is not wired up
to break the protocol that badly.
This also includes tweaks to 'qemu-nbd --list' to show when a server
is advertising the capability, and to the testsuite to reflect the
addition to that output.
Of note: qemu will always advertise the new feature bit during
NBD_OPT_INFO if extended headers have alreay been negotiated
(regardless of whether any NBD_OPT_SET_META_CONTEXT negotiation has
occurred); but for NBD_OPT_GO, qemu only advertises the feature if
block status is also enabled (that is, if the client does not
negotiate any contexts, then NBD_CMD_BLOCK_STATUS cannot be used, so
the feature is not advertised).
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230925192229.3186470-26-eblake@redhat.com>
[eblake: fix logic to reject unnegotiated contexts]
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
The next commit will add support for the optional extension
NBD_CMD_FLAG_PAYLOAD during NBD_CMD_BLOCK_STATUS, where the client can
request that the server only return a subset of negotiated contexts,
rather than all contexts. To make that task easier, this patch
populates the list of contexts to return on a per-command basis (for
now, identical to the full set of negotiated contexts).
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230925192229.3186470-25-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
|
|
Peform several minor refactorings of how the list of negotiated meta
contexts is managed, to make upcoming patches easier: Promote the
internal type NBDExportMetaContexts to the public opaque type
NBDMetaContexts, and mark exp const. Use a shorter member name in
NBDClient. Hoist calls to nbd_check_meta_context() earlier in their
callers, as the number of negotiated contexts may impact the flags
exposed in regards to an export, which in turn requires a new
parameter. Drop a redundant parameter to nbd_negotiate_meta_queries.
No semantic change intended on the success path; on the failure path,
dropping context in nbd_check_meta_export even when reporting an error
is safer.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20230925192229.3186470-24-eblake@redhat.com>
|
|
Time to start supporting clients that request extended headers. Now
we can finally reach the code added across several previous patches.
Even though the NBD spec has been altered to allow us to accept
NBD_CMD_READ larger than the max payload size (provided our response
is a hole or broken up over more than one data chunk), we are not
planning to take advantage of that, and continue to cap NBD_CMD_READ
to 32M regardless of header size.
For NBD_CMD_WRITE_ZEROES and NBD_CMD_TRIM, the block layer already
supports 64-bit operations without any effort on our part. For
NBD_CMD_BLOCK_STATUS, the client's length is a hint, and the previous
patch took care of implementing the required
NBD_REPLY_TYPE_BLOCK_STATUS_EXT.
We do not yet support clients that want to do request payload
filtering of NBD_CMD_BLOCK_STATUS; that will be added in later
patches, but is not essential for qemu as a client since qemu only
requests the single context base:allocation.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20230925192229.3186470-19-eblake@redhat.com>
|
|
The NBD spec states that if the client negotiates extended headers,
the server must avoid NBD_REPLY_TYPE_BLOCK_STATUS and instead use
NBD_REPLY_TYPE_BLOCK_STATUS_EXT which supports 64-bit lengths, even if
the reply does not need more than 32 bits. As of this patch,
client->mode is still never NBD_MODE_EXTENDED, so the code added here
does not take effect until the next patch enables negotiation.
For now, all metacontexts that we know how to export never populate
more than 32 bits of information, so we don't have to worry about
NBD_REP_ERR_EXT_HEADER_REQD or filtering during handshake, and we
always send all zeroes for the upper 32 bits of status during
NBD_CMD_BLOCK_STATUS.
Note that we previously had some interesting size-juggling on call
chains, such as:
nbd_co_send_block_status(uint32_t length)
-> blockstatus_to_extents(uint32_t bytes)
-> bdrv_block_status_above(bytes, &uint64_t num)
-> nbd_extent_array_add(uint64_t num)
-> store num in 32-bit length
But we were lucky that it never overflowed: bdrv_block_status_above
never sets num larger than bytes, and we had previously been capping
'bytes' at 32 bits (since the protocol does not allow sending a larger
request without extended headers). This patch adds some assertions
that ensure we continue to avoid overflowing 32 bits for a narrow
client, while fully utilizing 64-bits all the way through when the
client understands that. Even in 64-bit math, overflow is not an
issue, because all lengths are coming from the block layer, and we
know that the block layer does not support images larger than off_t
(if lengths were coming from the network, the story would be
different).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20230925192229.3186470-18-eblake@redhat.com>
|
|
Although extended mode is not yet enabled, once we do turn it on, we
need to reply with extended headers to all messages. Update the low
level entry points necessary so that all other callers automatically
get the right header based on the current mode.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20230925192229.3186470-17-eblake@redhat.com>
|
|
Although extended mode is not yet enabled, once we do turn it on, we
need to accept extended requests for all messages. Previous patches
have already taken care of supporting 64-bit lengths, now we just need
to read it off the wire.
Note that this implementation will block indefinitely on a buggy
client that sends a non-extended payload (that is, we try to read a
full packet before we ever check the magic number, but a client that
mistakenly sends a simple request after negotiating extended headers
doesn't send us enough bytes), but it's no different from any other
client that stops talking to us partway through a packet and thus not
worth coding around.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20230925192229.3186470-16-eblake@redhat.com>
|
|
Upcoming additions to support NBD 64-bit effect lengths allow for the
possibility to distinguish between payload length (capped at 32M) and
effect length (64 bits, although we generally assume 63 bits because
of off_t limitations). Without that extension, only the NBD_CMD_WRITE
request has a payload; but with the extension, it makes sense to allow
at least NBD_CMD_BLOCK_STATUS to have both a payload and effect length
in a future patch (where the payload is a limited-size struct that in
turn gives the real effect length as well as a subset of known ids for
which status is requested). Other future NBD commands may also have a
request payload, so the 64-bit extension introduces a new
NBD_CMD_FLAG_PAYLOAD_LEN that distinguishes between whether the header
length is a payload length or an effect length, rather than
hard-coding the decision based on the command.
According to the spec, a client should never send a command with a
payload without the negotiation phase proving such extension is
available. So in the unlikely event the bit is set or cleared
incorrectly, the client is already at fault; if the client then
provides the payload, we can gracefully consume it off the wire and
fail the command with NBD_EINVAL (subsequent checks for magic numbers
ensure we are still in sync), while if the client fails to send
payload we block waiting for it (basically deadlocking our connection
to the bad client, but not negatively impacting our ability to service
other clients, so not a security risk). Note that we do not support
the payload version of BLOCK_STATUS yet.
This patch also fixes a latent bug introduced in b2578459: once
request->len can be 64 bits, assigning it to a 32-bit payload_len can
cause wraparound to 0 which then sets req->complete prematurely;
thankfully, the bug was not possible back then (it takes this and
later patches to even allow request->len larger than 32 bits; and
since previously the only 'payload_len = request->len' assignment was
in NBD_CMD_WRITE which also sets check_length, which in turn rejects
lengths larger than 32M before relying on any possibly-truncated value
stored in payload_len).
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230925192229.3186470-15-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
[eblake: enhance comment on handling client error, fix type bug]
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Upcoming additions to support NBD 64-bit effect lengths will add a new
command flag NBD_CMD_FLAG_PAYLOAD_LEN that needs to be considered in
our sanity checks of the client's messages (that is, more than just
CMD_WRITE have the potential to carry a client payload when extended
headers are in effect). But before we can start to support that, it
is easier to first refactor the existing set of various if statements
over open-coded combinations of request->type to instead be a single
switch statement over all command types that sets witnesses, then
straight-line processing based on the witnesses. No semantic change
is intended.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230829175826.377251-24-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
|
|
Widen the length field of NBDRequest to 64-bits, although we can
assert that all current uses are still under 32 bits: either because
of NBD_MAX_BUFFER_SIZE which is even smaller (and where size_t can
still be appropriate, even on 32-bit platforms), or because nothing
ever puts us into NBD_MODE_EXTENDED yet (and while future patches will
allow larger transactions, the lengths in play here are still capped
at 32-bit). There are no semantic changes, other than a typo fix in a
couple of error messages.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230829175826.377251-23-eblake@redhat.com>
[eblake: fix assertion bug in nbd_co_send_simple_reply]
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
|
|
Add the constants and structs necessary for later patches to start
implementing the NBD_OPT_EXTENDED_HEADERS extension in both the client
and server, matching recent upstream nbd.git (through commit
e6f3b94a934). This patch does not change any existing behavior, but
merely sets the stage for upcoming patches.
This patch does not change the status quo that neither the client nor
server use a packed-struct representation for the request header.
While most of the patch adds new types, there is also some churn for
renaming the existing NBDExtent to NBDExtent32 to contrast it with
NBDExtent64, which I thought was a nicer name than NBDExtentExt.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20230829175826.377251-22-eblake@redhat.com>
|
|
The upcoming patches for 64-bit extensions requires various points in
the protocol to make decisions based on what was negotiated. While we
could easily add a 'bool extended_headers' alongside the existing
'bool structured_reply', this does not scale well if more modes are
added in the future. Better is to expose the mode enum added in the
recent commit bfe04d0a7d out to a wider use in the code base.
Where the code previously checked for structured_reply being set or
clear, it now prefers checking for an inequality; this works because
the nodes are in a continuum of increasing abilities, and allows us to
touch fewer places if we ever insert other modes in the middle of the
enum. There should be no semantic change in this patch.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230829175826.377251-20-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
|
|
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>
|
|
Externally, libnbd exposed the 64-bit opaque marker for each client
NBD packet as the "cookie", because it was less confusing when
contrasted with 'struct nbd_handle *' holding all libnbd state. It
also avoids confusion between the noun 'handle' as a way to identify a
packet and the verb 'handle' for reacting to things like signals.
Upstream NBD changed their spec to favor the name "cookie" based on
libnbd's recommendations[1], so we can do likewise.
[1] https://github.com/NetworkBlockDevice/nbd/commit/ca4392eb2b
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230608135653.2918540-6-eblake@redhat.com>
[eblake: typo fix]
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
|
|
Part of NBD's 64-bit headers extension involves passing the client's
requested offset back as part of the reply header (one reason it
stated for this change: converting absolute offsets stored in
NBD_REPLY_TYPE_OFFSET_DATA to relative offsets within the buffer is
easier if the absolute offset of the buffer is also available). This
is a refactoring patch to pass the full request around the reply
stack, rather than just the handle, so that later patches can then
access request->from when extended headers are active. Meanwhile,
this patch enables us to now assert that simple replies are only
attempted when appropriate, and otherwise has no semantic change.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20230608135653.2918540-5-eblake@redhat.com>
|
|
Upstream NBD now documents[1] an extension that supports 64-bit effect
lengths in requests. As part of that extension, the size of the reply
headers will change in order to permit a 64-bit length in the reply
for symmetry[2]. Additionally, where the reply header is currently 16
bytes for simple reply, and 20 bytes for structured reply; with the
extension enabled, there will only be one extended reply header, of 32
bytes, with both structured and extended modes sending identical
payloads for chunked replies.
Since we are already wired up to use iovecs, it is easiest to allow
for this change in header size by splitting each structured reply
across multiple iovecs, one for the header (which will become wider in
a future patch according to client negotiation), and the other(s) for
the chunk payload, and removing the header from the payload struct
definitions. Rename the affected functions with s/structured/chunk/
to make it obvious that the code will be reused in extended mode.
Interestingly, the client side code never utilized the packed types,
so only the server code needs to be updated.
[1] https://github.com/NetworkBlockDevice/nbd/blob/extension-ext-header/doc/proto.md
as of NBD commit e6f3b94a934
[2] Note that on the surface, this is because some future server might
permit a 4G+ NBD_CMD_READ and need to reply with that much data in one
transaction. But even though the extended reply length is widened to
64 bits, for now the NBD spec is clear that servers will not reply
with more than a maximum payload bounded by the 32-bit
NBD_INFO_BLOCK_SIZE field; allowing a client and server to mutually
agree to transactions larger than 4G would require yet another
extension.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230608135653.2918540-4-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
|
|
Mechanical change running Coccinelle spatch with content
generated from the qom-cast-macro-clean-cocci-gen.py added
in the previous commit.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20230601093452.38972-3-philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Thomas Huth <thuth@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>
|
|
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
exp->common.blk cannot be NULL, nbd_export_delete() is only called (through
a bottom half) from blk_exp_unref() and in turn that can only happen
after blk_exp_add() has asserted exp->blk != NULL.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Nagle's algorithm adds latency in order to reduce network packet
overhead on small packets. But when we are already using corking to
merge smaller packets into transactional requests, the extra delay
from TCP defaults just gets in the way (see recent commit bd2cd4a4).
For reference, qemu as an NBD client already requests TCP_NODELAY (see
nbd_connect() in nbd/client-connection.c); as does libnbd as a client
[1], and nbdkit as a server [2]. Furthermore, the NBD spec recommends
the use of TCP_NODELAY [3].
[1] https://gitlab.com/nbdkit/libnbd/-/blob/a48a1142/generator/states-connect.c#L39
[2] https://gitlab.com/nbdkit/nbdkit/-/blob/45b72f5b/server/sockets.c#L430
[3] https://github.com/NetworkBlockDevice/nbd/blob/master/doc/proto.md#protocol-phases
CC: Florian Westphal <fw@strlen.de>
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20230404004047.142086-1-eblake@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
|
|
qemu-nbd doesn't set TCP_NODELAY on the tcp socket.
Kernel waits for more data and avoids transmission of small packets.
Without TLS this is barely noticeable, but with TLS this really shows.
Booting a VM via qemu-nbd on localhost (with tls) takes more than
2 minutes on my system. tcpdump shows frequent wait periods, where no
packets get sent for a 40ms period.
Add explicit (un)corking when processing (and responding to) requests.
"TCP_CORK, &zero" after earlier "CORK, &one" will flush pending data.
VM Boot time:
main: no tls: 23s, with tls: 2m45s
patched: no tls: 14s, with tls: 15s
VM Boot time, qemu-nbd via network (same lan):
main: no tls: 18s, with tls: 1m50s
patched: no tls: 17s, with tls: 18s
Future optimization: if we could detect if there is another pending
request we could defer the uncork operation because more data would be
appended.
Signed-off-by: Florian Westphal <fw@strlen.de>
Message-Id: <20230324104720.2498-1-fw@strlen.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
We have two inclusion loops:
block/block.h
-> block/block-global-state.h
-> block/block-common.h
-> block/blockjob.h
-> block/block.h
block/block.h
-> block/block-io.h
-> block/block-common.h
-> block/blockjob.h
-> block/block.h
I believe these go back to Emanuele's reorganization of the block API,
merged a few months ago in commit d7e2fe4aac8.
Fortunately, breaking them is merely a matter of deleting unnecessary
includes from headers, and adding them back in places where they are
now missing.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20221221133551.3967339-2-armbru@redhat.com>
|
|
Avoid mixing bdrv_* functions with blk_*, so create blk_* counterparts
for bdrv_block_status_above and bdrv_is_allocated_above.
Note that since blk_co_block_status_above only calls the g_c_w function
bdrv_common_block_status_above and is marked as coroutine_fn, call
directly bdrv_co_common_block_status_above() to avoid using a g_c_w.
Same applies to blk_co_is_allocated_above.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20221128142337.657646-5-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
These functions end up calling bdrv_*() implemented as generated_co_wrapper
functions.
In addition, they also happen to be always called in coroutine context,
meaning all callers are coroutine_fn.
This means that the g_c_w function will enter the qemu_in_coroutine()
case and eventually suspend (or in other words call qemu_coroutine_yield()).
Therefore we can mark such functions coroutine_fn too.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20221128142337.657646-4-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
block-export-add argument @name defaults to the value of argument
@node-name.
nbd_export_create() implements this by copying @node_name to @name.
It leaves @has_node_name false, violating the "has_node_name ==
!!node_name" invariant. Unclean. Falls apart when we elide
@has_node_name (next commit): then QAPI frees the same value twice,
once for @node_name and once @name. iotest 307 duly explodes.
Goes back to commit c62d24e906 "blockdev-nbd: Boxed argument type for
nbd-server-add" (v5.0.0). Got moved from qmp_nbd_server_add() to
nbd_export_create() (commit 56ee86261e), then copied back (commit
b6076afcab). Commit 8675cbd68b "nbd: Utilize QAPI_CLONE for type
conversion" (v5.2.0) cleaned up the copy in qmp_nbd_server_add()
noting
Second, our assignment to arg->name is fishy: the generated QAPI code
for qapi_free_NbdServerAddOptions does not visit arg->name if
arg->has_name is false, but if it DID visit it, we would have
introduced a double-free situation when arg is finally freed.
Exactly. However, the copy in nbd_export_create() remained dirty.
Clean it up. Since the value stored in member @name is not actually
used outside this function, use a local variable instead of modifying
the QAPI object.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Cc: Eric Blake <eblake@redhat.com>
Cc: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Cc: qemu-block@nongnu.org
Message-Id: <20221104160712.3005652-10-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
|
|
Swap 'buf' and 'bytes' around for consistency with
blk_co_{pread,pwrite}(), and in preparation to implement these functions
using generated_co_wrapper.
Callers were updated using this Coccinelle script:
@@ expression blk, offset, buf, bytes, flags; @@
- blk_pread(blk, offset, buf, bytes, flags)
+ blk_pread(blk, offset, bytes, buf, flags)
@@ expression blk, offset, buf, bytes, flags; @@
- blk_pwrite(blk, offset, buf, bytes, flags)
+ blk_pwrite(blk, offset, bytes, buf, flags)
It had no effect on hw/block/nand.c, presumably due to the #if, so that
file was updated manually.
Overly-long lines were then fixed by hand.
Signed-off-by: Alberto Faria <afaria@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220705161527.1054072-4-afaria@redhat.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
|
|
For consistency with other I/O functions, and in preparation to
implement it using generated_co_wrapper.
Callers were updated using this Coccinelle script:
@@ expression blk, offset, buf, bytes; @@
- blk_pread(blk, offset, buf, bytes)
+ blk_pread(blk, offset, buf, bytes, 0)
It had no effect on hw/block/nand.c, presumably due to the #if, so that
file was updated manually.
Overly-long lines were then fixed by hand.
Signed-off-by: Alberto Faria <afaria@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220705161527.1054072-3-afaria@redhat.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
|
|
According to the NBD spec, a server that advertises
NBD_FLAG_CAN_MULTI_CONN promises that multiple client connections will
not see any cache inconsistencies: when properly separated by a single
flush, actions performed by one client will be visible to another
client, regardless of which client did the flush.
We always satisfy these conditions in qemu - even when we support
multiple clients, ALL clients go through a single point of reference
into the block layer, with no local caching. The effect of one client
is instantly visible to the next client. Even if our backend were a
network device, we argue that any multi-path caching effects that
would cause inconsistencies in back-to-back actions not seeing the
effect of previous actions would be a bug in that backend, and not the
fault of caching in qemu. As such, it is safe to unconditionally
advertise CAN_MULTI_CONN for any qemu NBD server situation that
supports parallel clients.
Note, however, that we don't want to advertise CAN_MULTI_CONN when we
know that a second client cannot connect (for historical reasons,
qemu-nbd defaults to a single connection while nbd-server-add and QMP
commands default to unlimited connections; but we already have
existing means to let either style of NBD server creation alter those
defaults). This is visible by no longer advertising MULTI_CONN for
'qemu-nbd -r' without -e, as in the iotest nbd-qemu-allocation.
The harder part of this patch is setting up an iotest to demonstrate
behavior of multiple NBD clients to a single server. It might be
possible with parallel qemu-io processes, but I found it easier to do
in python with the help of libnbd, and help from Nir and Vladimir in
writing the test.
Signed-off-by: Eric Blake <eblake@redhat.com>
Suggested-by: Nir Soffer <nsoffer@redhat.com>
Suggested-by: Vladimir Sementsov-Ogievskiy <v.sementsov-og@mail.ru>
Message-Id: <20220512004924.417153-3-eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
Hi all! Current logic of relying on search through backing chain is not
safe neither convenient.
Sometimes it leads to necessity of extra bitmap copying. Also, we are
going to add "snapshot-access" driver, to access some snapshot state
through NBD. And this driver is not formally a filter, and of course
it's not a COW format driver. So, searching through backing chain will
not work. Instead of widening the workaround of bitmap searching, let's
extend the interface so that user can select bitmap precisely.
Note, that checking for bitmap active status is not copied to the new
API, I don't see a reason for it, user should understand the risks. And
anyway, bitmap from other node is unrelated to this export being
read-only or read-write.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@openvz.org>
Message-Id: <20220314213226.362217-3-v.sementsov-og@mail.ru>
[eblake: Adjust S-o-b to Vladimir's new email, with permission]
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
The macro doesn't need it.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
|
|
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>
|
|
staging
nbd patches for 2022-03-07
- Dan Berrange: Allow qemu-nbd to support TLS over Unix sockets
- Eric Blake: Minor cleanups related to 64-bit block operations
# gpg: Signature made Tue 08 Mar 2022 01:41:35 GMT
# gpg: using RSA key 71C2CC22B1C4602927D2F3AAA7A16B4A2527436A
# gpg: Good signature from "Eric Blake <eblake@redhat.com>" [full]
# gpg: aka "Eric Blake (Free Software Programmer) <ebb9@byu.net>" [full]
# gpg: aka "[jpeg image of size 6874]" [full]
# Primary key fingerprint: 71C2 CC22 B1C4 6029 27D2 F3AA A7A1 6B4A 2527 436A
* remotes/ericb/tags/pull-nbd-2022-03-07:
qemu-io: Allow larger write zeroes under no fallback
qemu-io: Utilize 64-bit status during map
nbd/server: Minor cleanups
tests/qemu-iotests: validate NBD TLS with UNIX sockets and PSK
tests/qemu-iotests: validate NBD TLS with UNIX sockets
tests/qemu-iotests: validate NBD TLS with hostname mismatch
tests/qemu-iotests: convert NBD TLS test to use standard filters
tests/qemu-iotests: introduce filter for qemu-nbd export list
tests/qemu-iotests: expand _filter_nbd rules
tests/qemu-iotests: add QEMU_IOTESTS_REGEN=1 to update reference file
block/nbd: don't restrict TLS usage to IP sockets
qemu-nbd: add --tls-hostname option for TLS certificate validation
block/nbd: support override of hostname for TLS certificate validation
block: pass desired TLS hostname through from block driver client
crypto: mandate a hostname when checking x509 creds on a client
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
|
|
Spelling fixes, grammar improvements and consistent spacing, noticed
while preparing other patches in this file.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20211203231539.3900865-2-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
|
|
Move the various memalign-related functions out of osdep.h and into
their own header, which we include only where they are used.
While we're doing this, add some brief documentation comments.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20220226180723.1706285-10-peter.maydell@linaro.org
|
|
NBDRequestData struct has unused QSIMPLEQ_ENTRY field. It seems that
this field exists since the first git commit and was never used.
Signed-off-by: Nir Soffer <nsoffer@redhat.com>
Message-Id: <20220111194313.581486-1-nsoffer@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Fixes: d9a73806 ("qemu-nbd: introduce NBDRequest", v1.1)
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Now that the block layer supports 64-bit operations (see commit
2800637a and friends, new to v6.2), we no longer have to self-fragment
requests larger than 2G, reverting the workaround added in 890cbccb08
("nbd: Fix large trim/zero requests", v5.1.0).
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20211117170230.1128262-3-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
|
|
When a client disconnects abruptly, but did not have any pending
requests (for example, when using nbdsh without calling h.shutdown),
we used to output the following message:
$ qemu-nbd -f raw file
$ nbdsh -u 'nbd://localhost:10809' -c 'h.trim(1,0)'
qemu-nbd: Disconnect client, due to: Failed to read request: Unexpected end-of-file before all bytes were read
Then in commit f148ae7, we refactored nbd_receive_request() to use
nbd_read_eof(); when this returns 0, we regressed into tracing
uninitialized memory (if tracing is enabled) and reporting a
less-specific:
qemu-nbd: Disconnect client, due to: Request handling failed in intermediate state
Note that with Unix sockets, we have yet another error message,
unchanged by the 6.0 regression:
$ qemu-nbd -k /tmp/sock -f raw file
$ nbdsh -u 'nbd+unix:///?socket=/tmp/sock' -c 'h.trim(1,0)'
qemu-nbd: Disconnect client, due to: Failed to send reply: Unable to write to socket: Broken pipe
But in all cases, the error message goes away if the client performs a
soft shutdown by using NBD_CMD_DISC, rather than a hard shutdown by
abrupt disconnect:
$ nbdsh -u 'nbd://localhost:10809' -c 'h.trim(1,0)' -c 'h.shutdown()'
This patch fixes things to avoid uninitialized memory, and in general
avoids warning about a client that does a hard shutdown when not in
the middle of a packet. A client that aborts mid-request, or which
does not read the full server's reply, can still result in warnings,
but those are indeed much more unusual situations.
CC: qemu-stable@nongnu.org
Fixes: f148ae7d36 ("nbd/server: Quiesce coroutines on context switch", v6.0.0)
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
[eblake: defer unrelated typo fixes to later patch]
Message-Id: <20211117170230.1128262-2-eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
clang's sanitizer is picky: memset(NULL, x, 0) is technically
undefined behavior, even though no sane implementation of memset()
deferences the NULL. Caught by the nbd-qemu-allocation iotest.
The alternative to checking before each memset is to instead force an
allocation of 1 element instead of g_new0(type, 0)'s behavior of
returning NULL for a 0-length array.
Reported-by: Peter Maydell <peter.maydell@linaro.org>
Fixes: 3b1f244c59 (nbd: Allow export of multiple bitmaps for one device)
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20211115223943.626416-1-eblake@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
|
|
The NBD protocol just relaxed the requirements on
NBD_OPT_LIST_META_CONTEXT:
https://github.com/NetworkBlockDevice/nbd/commit/13a4e33a87
Since listing is not stateful (unlike SET_META_CONTEXT), we don't care
if a client asks for meta contexts without first requesting structured
replies. Well-behaved clients will still ask for structured reply
first (if for no other reason than for back-compat to older servers),
but that's no reason to avoid this change.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210907173505.1499709-1-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
|
|
From clang-13:
nbd/server.c:976:22: error: variable 'bitmaps' set but not used \
[-Werror,-Wunused-but-set-variable]
which is incorrect; see //bugs.llvm.org/show_bug.cgi?id=3888.
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
|
|
Before switching between AioContexts we need to make sure that we're
fully quiesced ("nb_requests == 0" for every client) when entering the
drained section.
To do this, we set "quiescing = true" for every client on
".drained_begin" to prevent new coroutines from being created, and
check if "nb_requests == 0" on ".drained_poll". Finally, once we're
exiting the drained section, on ".drained_end" we set "quiescing =
false" and call "nbd_client_receive_next_request()" to resume the
processing of new requests.
With these changes, "blk_aio_attach()" and "blk_aio_detach()" can be
reverted to be as simple as they were before f148ae7d36.
RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=1960137
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Sergio Lopez <slp@redhat.com>
Message-Id: <20210602060552.17433-3-slp@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
|
|
When querying image extents for raw image, qemu-nbd reports holes as
zero:
$ qemu-nbd -t -r -f raw empty-6g.raw
$ qemu-img map --output json nbd://localhost
[{ "start": 0, "length": 6442450944, "depth": 0, "zero": true, "data": true, "offset": 0}]
$ qemu-img map --output json empty-6g.raw
[{ "start": 0, "length": 6442450944, "depth": 0, "zero": true, "data": false, "offset": 0}]
Turns out that qemu-img map reports a hole based on BDRV_BLOCK_DATA, but
nbd server reports a hole based on BDRV_BLOCK_ALLOCATED.
The NBD protocol says:
NBD_STATE_HOLE (bit 0): if set, the block represents a hole (and
future writes to that area may cause fragmentation or encounter an
NBD_ENOSPC error); if clear, the block is allocated or the server
could not otherwise determine its status.
qemu-img manual says:
whether the sectors contain actual data or not (boolean field data;
if false, the sectors are either unallocated or stored as
optimized all-zero clusters);
To me, data=false looks compatible with NBD_STATE_HOLE. From user point
of view, getting same results from qemu-nbd and qemu-img is more
important than being more correct about allocation status.
Changing nbd server to report holes using BDRV_BLOCK_DATA makes qemu-nbd
results compatible with qemu-img map:
$ qemu-img map --output json nbd://localhost
[{ "start": 0, "length": 6442450944, "depth": 0, "zero": true, "data": false, "offset": 0}]
Signed-off-by: Nir Soffer <nsoffer@redhat.com>
Message-Id: <20210219160752.1826830-1-nsoffer@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
When switching between AIO contexts we need to me make sure that both
recv_coroutine and send_coroutine are not scheduled to run. Otherwise,
QEMU may crash while attaching the new context with an error like
this one:
aio_co_schedule: Co-routine was already scheduled in 'aio_co_schedule'
To achieve this we need a local implementation of
'qio_channel_readv_all_eof' named 'nbd_read_eof' (a trick already done
by 'nbd/client.c') that allows us to interrupt the operation and to
know when recv_coroutine is yielding.
With this in place, we delegate detaching the AIO context to the
owning context with a BH ('nbd_aio_detach_bh') scheduled using
'aio_wait_bh_oneshot'. This BH signals that we need to quiesce the
channel by setting 'client->quiescing' to 'true', and either waits for
the coroutine to finish using AIO_WAIT_WHILE or, if it's yielding in
'nbd_read_eof', actively enters the coroutine to interrupt it.
RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=1900326
Signed-off-by: Sergio Lopez <slp@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201214170519.223781-4-slp@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Coverity noticed (CID 1436125) that we check the return value of
nbd_extent_array_add in most places, but not at the end of
bitmap_to_extents(). The return value exists to break loops before a
future iteration, so there is nothing to check if we are already done
iterating. Adding a cast to void, plus a comment why, pacifies
Coverity.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20201111163510.713855-1-eblake@redhat.com>
[eblake: Prefer cast to void over odd && usage]
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
|
