As nvme_create_queue_pair() is allowed to fail, replace the
alloc() calls by try_alloc() to avoid aborting QEMU.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200821195359.1285345-4-philmd@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Avoid further processing if TRACE_NVME_SUBMIT_COMMAND_RAW is
not enabled. This is an untested intend of performance optimization.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200821195359.1285345-3-philmd@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Use self-explicit SCALE_MS definition instead of magic value.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200821195359.1285345-2-philmd@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The NVM Express specification generally uses 'zeroes' and not 'zeros',
so let us align with it.
Cc: Fam Zheng <fam@euphon.net>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Minwoo Im <minwoo.im.dev@gmail.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Add missing fields in the Identify Controller and Identify Namespace
data structures to bring them in line with NVMe v1.3.
This also adds data structures and defines for SGL support which
requires a couple of trivial changes to the nvme block driver as well.
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Acked-by: Fam Zheng <fam@euphon.net>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Dmitry Fomichev <dmitry.fomichev@wdc.com>
Message-Id: <20200706061303.246057-2-its@irrelevant.dk>
QEMU block drivers are supposed to support aio_poll() from I/O
completion callback functions. This means completion processing must be
re-entrant.
The standard approach is to schedule a BH during completion processing
and cancel it at the end of processing. If aio_poll() is invoked by a
callback function then the BH will run. The BH continues the suspended
completion processing.
All of this means that request A's cb() can synchronously wait for
request B to complete. Previously the nvme block driver would hang
because it didn't process completions from nested aio_poll().
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Sergio Lopez <slp@redhat.com>
Message-id: 20200617132201.1832152-8-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Passing around both BDRVNVMeState and NVMeQueuePair is unwieldy. Reduce
the number of function arguments by keeping the BDRVNVMeState pointer in
NVMeQueuePair. This will come in handly when a BH is introduced in a
later patch and only one argument can be passed to it.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Sergio Lopez <slp@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200617132201.1832152-7-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
There are three issues with the current NVMeRequest->busy field:
1. The busy field is accidentally accessed outside q->lock when request
submission fails.
2. Waiters on free_req_queue are not woken when a request is returned
early due to submission failure.
2. Finding a free request involves scanning all requests. This makes
request submission O(n^2).
Switch to an O(1) freelist that is always accessed under the lock.
Also differentiate between NVME_QUEUE_SIZE, the actual SQ/CQ size, and
NVME_NUM_REQS, the number of usable requests. This makes the code
simpler than using NVME_QUEUE_SIZE everywhere and having to keep in mind
that one slot is reserved.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Sergio Lopez <slp@redhat.com>
Message-id: 20200617132201.1832152-5-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Do not access a CQE after incrementing q->cq.head and releasing q->lock.
It is unlikely that this causes problems in practice but it's a latent
bug.
The reason why it should be safe at the moment is that completion
processing is not re-entrant and the CQ doorbell isn't written until the
end of nvme_process_completion().
Make this change now because QEMU expects completion processing to be
re-entrant and later patches will do that.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Sergio Lopez <slp@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200617132201.1832152-4-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
A lot of CPU time is spent simply locking/unlocking q->lock during
polling. Check for completion outside the lock to make q->lock disappear
from the profile.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Sergio Lopez <slp@redhat.com>
Message-id: 20200617132201.1832152-2-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Compress two lines into a single line if immediate return statement is found.
It also remove variables progress, val, data, ret and sock
as they are no longer needed.
Remove space between function "mixer_load" and '(' to fix the
checkpatch.pl error:-
ERROR: space prohibited between function name and open parenthesis '('
Done using following coccinelle script:
@@
local idexpression ret;
expression e;
@@
-ret =
+return
e;
-return ret;
Signed-off-by: Simran Singhal <singhalsimran0@gmail.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200401165314.GA3213@simran-Inspiron-5558>
[lv: in handle_aiocb_write_zeroes_unmap() move "int ret" inside the #ifdef]
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Instead of checking the .bdrv_co_create_opts to see if we need the
fallback, just implement the .bdrv_co_create_opts in the drivers that
need it.
This way we don't break various places that need to know if the
underlying protocol/format really supports image creation, and this way
we still allow some drivers to not support image creation.
Fixes: fd17146cd9
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1816007
Note that technically this driver reverts the image creation fallback
for the vxhs driver since I don't have a means to test it, and IMHO it
is better to leave it not supported as it was prior to generic image
creation patches.
Also drop iscsi_create_opts which was left accidentally.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20200326011218.29230-3-mlevitsk@redhat.com>
Reviewed-by: Denis V. Lunev <den@openvz.org>
[mreitz: Fixed alignment, and moved bdrv_co_create_opts_simple() and
bdrv_create_opts_simple from block.h into block_int.h]
Signed-off-by: Max Reitz <mreitz@redhat.com>
Replay is capable of recording normal BH events, but sometimes
there are single use callbacks scheduled with aio_bh_schedule_oneshot
function. This patch enables recording and replaying such callbacks.
Block layer uses these events for calling the completion function.
Replaying these calls makes the execution deterministic.
Signed-off-by: Pavel Dovgalyuk <Pavel.Dovgaluk@ispras.ru>
Acked-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In my "build everything" tree, changing qemu/main-loop.h triggers a
recompile of some 5600 out of 6600 objects (not counting tests and
objects that don't depend on qemu/osdep.h). It includes block/aio.h,
which in turn includes qemu/event_notifier.h, qemu/notify.h,
qemu/processor.h, qemu/qsp.h, qemu/queue.h, qemu/thread-posix.h,
qemu/thread.h, qemu/timer.h, and a few more.
Include qemu/main-loop.h only where it's needed. Touching it now
recompiles only some 1700 objects. For block/aio.h and
qemu/event_notifier.h, these numbers drop from 5600 to 2800. For the
others, they shrink only slightly.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20190812052359.30071-21-armbru@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Linux does not support blocks greater than 4 kB anyway, so we might as
well limit blkshift to 12 and thus save us from some potential trouble.
Reported-by: Peter Maydell <peter.maydell@linaro.org>
Suggested-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190730114812.10493-1-mreitz@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Coverity: CID 1403771
Signed-off-by: Max Reitz <mreitz@redhat.com>
Completion entries are meant to be only read by the host and written by the device.
The driver is supposed to scan the completions from the last point where it left,
and until it sees a completion with non flipped phase bit.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190716163020.13383-4-mlevitsk@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Currently the driver hardcodes the sector size to 512,
and doesn't check the underlying device. Fix that.
Also fail if underlying nvme device is formatted with metadata
as this needs special support.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-id: 20190716163020.13383-3-mlevitsk@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Fix the math involving non standard doorbell stride
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190716163020.13383-2-mlevitsk@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
When creating the admin queue in nvme_init() the variable that
holds the number of queues created is modified before actual
queue creation. This is a problem because if creating the queue
fails then the variable is left in inconsistent state. This was
actually observed when I tried to hotplug a nvme disk. The
control got to nvme_file_open() which called nvme_init() which
failed and thus nvme_close() was called which in turn called
nvme_free_queue_pair() with queue being NULL. This lead to an
instant crash:
#0 0x000055d9507ec211 in nvme_free_queue_pair (bs=0x55d952ddb880, q=0x0) at block/nvme.c:164
#1 0x000055d9507ee180 in nvme_close (bs=0x55d952ddb880) at block/nvme.c:729
#2 0x000055d9507ee3d5 in nvme_file_open (bs=0x55d952ddb880, options=0x55d952bb1410, flags=147456, errp=0x7ffd8e19e200) at block/nvme.c:781
#3 0x000055d9507629f3 in bdrv_open_driver (bs=0x55d952ddb880, drv=0x55d95109c1e0 <bdrv_nvme>, node_name=0x0, options=0x55d952bb1410, open_flags=147456, errp=0x7ffd8e19e310) at block.c:1291
#4 0x000055d9507633d6 in bdrv_open_common (bs=0x55d952ddb880, file=0x0, options=0x55d952bb1410, errp=0x7ffd8e19e310) at block.c:1551
#5 0x000055d950766881 in bdrv_open_inherit (filename=0x0, reference=0x0, options=0x55d952bb1410, flags=32768, parent=0x55d9538ce420, child_role=0x55d950eaade0 <child_file>, errp=0x7ffd8e19e510) at block.c:3063
#6 0x000055d950765ae4 in bdrv_open_child_bs (filename=0x0, options=0x55d9541cdff0, bdref_key=0x55d950af33aa "file", parent=0x55d9538ce420, child_role=0x55d950eaade0 <child_file>, allow_none=true, errp=0x7ffd8e19e510) at block.c:2712
#7 0x000055d950766633 in bdrv_open_inherit (filename=0x0, reference=0x0, options=0x55d9541cdff0, flags=0, parent=0x0, child_role=0x0, errp=0x7ffd8e19e908) at block.c:3011
#8 0x000055d950766dba in bdrv_open (filename=0x0, reference=0x0, options=0x55d953d00390, flags=0, errp=0x7ffd8e19e908) at block.c:3156
#9 0x000055d9507cb635 in blk_new_open (filename=0x0, reference=0x0, options=0x55d953d00390, flags=0, errp=0x7ffd8e19e908) at block/block-backend.c:389
#10 0x000055d950465ec5 in blockdev_init (file=0x0, bs_opts=0x55d953d00390, errp=0x7ffd8e19e908) at blockdev.c:602
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Message-id: 927aae40b617ba7d4b6c7ffe74e6d7a2595f8e86.1562770546.git.mprivozn@redhat.com
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Currently, nvme's bdrv_refresh_filename() is an exact copy of null's
implementation. However, for null, "null-co://" and "null-aio://" are
indeed valid filenames -- for nvme, they are not, as a device address is
still required.
The correct implementation should generate a filename of the form
"nvme://[PCI address]/[namespace]" (as the comment above
nvme_parse_filename() describes).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-id: 20190201192935.18394-27-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Currently, BlockDriver.bdrv_refresh_filename() is supposed to both
refresh the filename (BDS.exact_filename) and set BDS.full_open_options.
Now that we have generic code in the central bdrv_refresh_filename() for
creating BDS.full_open_options, we can drop the latter part from all
BlockDriver.bdrv_refresh_filename() implementations.
This also means that we can drop all of the existing default code for
this from the global bdrv_refresh_filename() itself.
Furthermore, we now have to call BlockDriver.bdrv_refresh_filename()
after having set BDS.full_open_options, because the block driver's
implementation should now be allowed to depend on BDS.full_open_options
being set correctly.
Finally, with this patch we can drop the @options parameter from
BlockDriver.bdrv_refresh_filename(); also, add a comment on this
function's purpose in block/block_int.h while touching its interface.
This completely obsoletes blklogwrite's implementation of
.bdrv_refresh_filename().
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20190201192935.18394-25-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
This new field can be set by block drivers to list the runtime options
they accept that may influence the contents of the respective BDS. As of
a follow-up patch, this list will be used by the common
bdrv_refresh_filename() implementation to decide which options to put
into BDS.full_open_options (and consequently whether a JSON filename has
to be created), thus freeing the drivers of having to implement that
logic themselves.
Additionally, this patch adds the field to all of the block drivers that
need it and sets it accordingly.
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-id: 20190201192935.18394-22-mreitz@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
The QEMU_PACKED is causing a compiler warning/error with GCC 9:
CC block/nvme.o
block/nvme.c: In function ‘nvme_create_queue_pair’:
block/nvme.c:209:22: error: taking address of packed member of
‘struct <anonymous>’ may result in an unaligned pointer value
[-Werror=address-of-packed-member]
209 | q->sq.doorbell = &s->regs->doorbells[idx * 2 * s->doorbell_scale];
All members of the struct are naturally aligned, so there should
not be the need for QEMU_PACKED here, and the following QEMU_BUILD_BUG_ON
also ensures that there is no padding. Thus simply remove the QEMU_PACKED
here.
Buglink: https://bugs.launchpad.net/qemu/+bug/1817525
Reported-by: Satheesh Rajendran <sathnaga@linux.vnet.ibm.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In the block layer, synchronous APIs are often implemented by creating a
coroutine that calls the asynchronous coroutine-based implementation and
then waiting for completion with BDRV_POLL_WHILE().
For this to work with iothreads (more specifically, when the synchronous
API is called in a thread that is not the home thread of the block
device, so that the coroutine will run in a different thread), we must
make sure to call aio_wait_kick() at the end of the operation. Many
places are missing this, so that BDRV_POLL_WHILE() keeps hanging even if
the condition has long become false.
Note that bdrv_dec_in_flight() involves an aio_wait_kick() call. This
corresponds to the BDRV_POLL_WHILE() in the drain functions, but it is
generally not enough for most other operations because they haven't set
the return value in the coroutine entry stub yet. To avoid race
conditions there, we need to kick after setting the return value.
The race window is small enough that the problem doesn't usually surface
in the common path. However, it does surface and causes easily
reproducible hangs if the operation can return early before even calling
bdrv_inc/dec_in_flight, which many of them do (trivial error or no-op
success paths).
The bug in bdrv_truncate(), bdrv_check() and bdrv_invalidate_cache() is
slightly different: These functions even neglected to schedule the
coroutine in the home thread of the node. This avoids the hang, but is
obviously wrong, too. Fix those to schedule the coroutine in the right
AioContext in addition to adding aio_wait_kick() calls.
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
When the IO size is larger than 2 pages, we move the the pointer one by
one in the pagelist, this is inefficient.
This is a simple benchmark result:
Before:
$ qemu-io -c 'write 0 1G' nvme://0000:00:04.0/1
wrote 1073741824/1073741824 bytes at offset 0
1 GiB, 1 ops; 0:00:02.41 (424.504 MiB/sec and 0.4146 ops/sec)
$ qemu-io -c 'read 0 1G' nvme://0000:00:04.0/1
read 1073741824/1073741824 bytes at offset 0
1 GiB, 1 ops; 0:00:02.03 (503.055 MiB/sec and 0.4913 ops/sec)
After:
$ qemu-io -c 'write 0 1G' nvme://0000:00:04.0/1
wrote 1073741824/1073741824 bytes at offset 0
1 GiB, 1 ops; 0:00:02.17 (471.517 MiB/sec and 0.4605 ops/sec)
$ qemu-io -c 'read 0 1G' nvme://0000:00:04.0/1
read 1073741824/1073741824 bytes at offset 0
1 GiB, 1 ops; 0:00:01.94 (526.770 MiB/sec and 0.5144 ops/sec)
Signed-off-by: Li Feng <lifeng1519@gmail.com>
Message-Id: <20181101103807.25862-1-lifeng1519@gmail.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
nvme_poll_queues is already protected by q->lock, and
AIO callbacks are invoked outside the AioContext lock.
So remove the acquire/release pair in nvme_handle_event.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20180814062739.19640-1-pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
bdrv_io_plug/bdrv_io_unplug take care of keeping a nesting count,
so change s->plugged to just a bool.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20180813144320.12382-2-pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
It is wrong to leave this field as 1, as nvme_close() called in the
error handling code in nvme_file_open() will use it and try to free
s->queues again.
Another problem is the cleaning ups are duplicated between the fail*
labels of nvme_init() and nvme_file_open(), which calls nvme_close().
A third problem is nvme_close() misses g_free() and
event_notifier_cleanup().
Fix all of them.
Cc: qemu-stable@nongnu.org
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20180712025420.4932-1-famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
For convenience and clarity, make it possible to call qobject_ref() at
the time when the reference is associated with a variable, or
argument, by making qobject_ref() return the same pointer as given.
Use that to simplify the callers.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180419150145.24795-5-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Useless change to qobject_ref_impl() dropped, commit message improved
slightly]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Now that we can safely call QOBJECT() on QObject * as well as its
subtypes, we can have macros qobject_ref() / qobject_unref() that work
everywhere instead of having to use QINCREF() / QDECREF() for QObject
and qobject_incref() / qobject_decref() for its subtypes.
The replacement is mechanical, except I broke a long line, and added a
cast in monitor_qmp_cleanup_req_queue_locked(). Unlike
qobject_decref(), qobject_unref() doesn't accept void *.
Note that the new macros evaluate their argument exactly once, thus no
need to shout them.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180419150145.24795-4-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Rebased, semantic conflict resolved, commit message improved]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Commit bdd6a90 has a bug: drivers should never directly set
BDRV_BLOCK_ALLOCATED, but only io.c should do that (as needed).
Instead, drivers should report BDRV_BLOCK_DATA if it knows that
data comes from this BDS.
But let's look at the bigger picture: semantically, the nvme
driver is similar to the nbd, null, and raw drivers (no backing
file, all data comes from this BDS). But while two of those
other drivers have to supply the callback (null because it can
special-case BDRV_BLOCK_ZERO, raw because it can special-case
a different offset), in this case the block layer defaults are
good enough without the callback at all (similar to nbd).
So, fix the bug by deletion ;)
Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
1) string not null terminated in sysfs_find_group_file
2) NULL pointer dereference and dead local variable in nvme_init.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20180213015240.9352-1-famz@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
qemu-common.h includes qemu/option.h, but most places that include the
former don't actually need the latter. Drop the include, and add it
to the places that actually need it.
While there, drop superfluous includes of both headers, and
separate #include from file comment with a blank line.
This cleanup makes the number of objects depending on qemu/option.h
drop from 4545 (out of 4743) to 284 in my "build everything" tree.
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20180201111846.21846-20-armbru@redhat.com>
[Semantic conflict with commit bdd6a90a9e in block/nvme.c resolved]
Forward these two calls to the IOVA manager.
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20180116060901.17413-6-famz@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
This is a new protocol driver that exclusively opens a host NVMe
controller through VFIO. It achieves better latency than linux-aio by
completely bypassing host kernel vfs/block layer.
$rw-$bs-$iodepth linux-aio nvme://
----------------------------------------
randread-4k-1 10.5k 21.6k
randread-512k-1 745 1591
randwrite-4k-1 30.7k 37.0k
randwrite-512k-1 1945 1980
(unit: IOPS)
The driver also integrates with the polling mechanism of iothread.
This patch is co-authored by Paolo and me.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20180116060901.17413-4-famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>