Handle IOs with metadata being transferred in a separate buffer if the
device underneath supports it.
Change-Id: I38887a24d2aad51f674a840367b5dcdeda2d5a8b
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/451467
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
This patch adds *_with_md family of functions allowing for IO with
metadata being transferred in a separate buffer.
Change-Id: I842d5a00a532cf5d0b0f0738535ea46903674140
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/451465
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
There is no need to keep this functionality in separate
files.
Change-Id: Ie998324cbcfcdc384f912fe36124b082774c0dc8
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456475
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
vring notification mechanism is transport-specific. At present, vhost
dataplane code in `lib/vhost/vhost.c` triggers guest notifications with
`eventfd_write()` system call. But this is an AF_UNIX specific
notification mechanism. This patch replaces `eventfd_write()` with the
existing generic `rte_vhost_vring_call()` function that is part of
DPDK's librte_vhost public API.
`rte_vhost_vring_call()` takes a vring_idx as an argument to associate
the `struct spdk_vhost_virtqueue` instance with the relevant `struct
vhost_virtqueue` instance. We introduce a new `vring_idx` field in
`struct spdk_vhost_virtqueue` to enable this association. This field is
initialized in `start_device()`. In addition, a stub for
`rte_vhost_vring_call()` is added in the vhost unit test file.
SPDK's internal `rte_vhost` copy will not be updated in order to support
the virtio-vhost-user transport. However, an `rte_vhost_vring_call()`
function is introduced in SPDK's `rte_vhost` in order to have a solid
API. This function is just a wrapper of `eventfd_write()`.
Change-Id: Ic93e25cd3f06e92f04766521bc850f1ee80b8ec8
Signed-off-by: Nikos Dragazis <ndragazis@arrikto.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/454373
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Request may be submitted several times via nvme_qpair_submit_request
function, such as request in queued_req queue being re-submitted.
With enabling timeout feature, nvme_qpair_submit_request compares
request->submit_tick to zero to check if this is the first submission
for this request. If true, record submit_tick for this reuqest.
So request->submit_tick needs to be set zero in allocation.
Change-Id: Ie3f420aa337802c5ad3962c3fdcd680dec1ccdcb
Signed-off-by: lorneli <lorneli@163.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456328
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This assert is not needed anymore.
Change-Id: I7bc85c980c2e120b105b35763b9bea5960564acc
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456590
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Keep an array of iovecs tied to each IO. Internal IOs can have
iov_cnt > 1 without having to allocate additional memory. User's iovec
can now be modified too (e.g. when splitting the request).
Change-Id: Iec73c6dad59acdc331a1461fd7feac085138a8de
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455525
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Replaced single IO pointer with a queue. This allows multiple requests
to be queued when the writes cannot be scheduled.
Change-Id: I0cde50e7378108a6aab4ef6fe601cb854244f5d5
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455524
Reviewed-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
On the user IO path, the flow always follows the same procedure:
allocate ftl_io and initialize it using ftl_io_user_init. There's no
point in separating these two actions.
Change-Id: I578347ccc7c85e5945368dd6bf76c58985af3939
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455523
Reviewed-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
There's no point in synchronously returning an error from ftl_io_write
/ ftl_io_read as they're also reported in the IO's completion callback.
All errors are now reported through io->status.
Change-Id: I4adf4e13221b63715625276042e1172cd63b8f9b
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455521
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Select the method of choosing next PPA to read inside ftl_submit_read,
as it can be deduced from other arguments.
Change-Id: I6cbd37c2c6d7f073bf735491b4960d72e327c4e6
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455520
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Some of the internal IOs split the contiguous data buffer into multiple
iovecs to separate the requests. It's unnecessary and can be replaced
by checking the position within the data buffer.
Change-Id: I9255ea0072fee6c4e6a7dca21e4008780bc45610
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455519
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Advance parent's IO position when children positions are advanced.
There's no need to call ftl_io_advance twice for the same request
anymore.
Change-Id: I1f7f04a3a83575b11e7bf0b13c0c8f3bf98b5522
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455518
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
This is a small optimization.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ib593908d3aeb17aac55be06b8e3be42e28a23061
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456268
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Existing code in spdk_bdev_write_zeroes_blocks() will call spdk_bdev_free_io()
for the error case, which will cause assertion because the bdev_io isn't
submitted to the backend yet, so we will check the condtion first to
avoid the error case.
Change-Id: If27d78217f709a3315e74c00869d345abd6b9a69
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453491
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
We were printing out the wrong value here.
Change-Id: I7b5f4eaca41317a69167ad5413a1b1913e9c0842
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456278
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This module simply sits on another virtual bdev
and adds a simulated average and p99 latency to that drive.
Change-Id: Ie9fc91e27585fd0636cb7dc845cb41744bf24625
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453594
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Using AVX512 or AVX2 ends up being a small pessimization.
I think AVX works better for copies when there are
multiple cachelines to copy. I see a 2-3% improvement
in high IOPs benchmarks when reverting to SSE.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I3d70a1e359e98cec2a9da41ccf9af2de9baa5868
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456247
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Profiling showed these weren't getting inlined - so add
the inline keyword to make sure it happens. This helps
improve performance a bit.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia86edccc9163258efdcddcce6989a71fb180caf6
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456099
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
At 10M IO/s, we see a lot of CPU cycles wasted getting
the next tracker into cache. If we only get one
completion at a time, this is unavoidable, but when
there are multiple completions pending, we can prefetch
the second tracker while processing the completion for
the first.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9de702bee3719e4494eec6f05b09be3672f1e0ac
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456097
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Prior merge contained all of the code EXCEPT for the user-callable function.
Signed-off-by: James Bergsten <jamesx.bergsten@intel.com>
Change-Id: I1cb7105ab85ffae8ed4f600261fed86c9c778893
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456282
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
All the SCSI commands have two states: allowed and conflict,
based on different reservation types and I_T nexus, the SCSI
module will check each commands before sending to the backend
device.
Change-Id: Ib05cece1c237873360f85c68d139362200d2d47e
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/436097
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Purpose: Make the variable definition consistent
with the same variable in the target side.
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: Ibc4ff92b6346f0a1ad803dcb79d041289f5648b2
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455807
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
QAT driver will not load with anything less than 128, may need to
adjust after performance testing.
Change-Id: I4949e6d97e7ed997d14b907dac00cc288dcc331b
Signed-off-by: paul luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456069
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
mbufs are freed with a single call to the first mbuf in the chain
unless we hit an error before we chain them. So free them correctly
when chained and loop through the array one at a time and free them
that way if not.
Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I80ebb86bc2147b42db89d24f8e985cacf7d4bceb
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455019
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
We need to move sub_reads++ statement earlier. Otherwise if
the read fails, the sem_wait(&channel->sem) call number
is not correct.
Change-Id: I05281a74bef78e4f80ce8d202b47849d1e6c2009
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453788
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Signed-off-by: Orden Smith <orden.e.smith@intel.com>
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Change-Id: I4a3c4d1ca8d0b18201edf99a67a3d75ca7ab8153
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/449499
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The purpose of is patch is to track whether there is
req allocation failure for those file metadata related
operation in order to find the bug.
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: I171a5b7fca214675fb1356ca650a01eeb20b4c8c
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/454829
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
_dif_sgl_fast_forward covers _dif_sgl_advance. Hence this patch
merges two into _dif_sgl_fast_forward and renames it to
_dif_sgl_advance.
Change-Id: I4949f3028b08cec8df8c9037c983d840d0a1eaaa
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456184
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This patch add support for VMD driver object.
New PCI device ID for VMD device was added.
Change-Id: I47bd8772a15ad370a14b7cc9460a177c91e6dd6a
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Signed-off-by: Orden Smith <orden.e.smith@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455545
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie32a1bb144c239b923b5cbb9e608a7dfc9c05208
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456076
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Maciej Szwed <maciej.szwed@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
In Fedora release 28, plug in nvme device and run setup.sh,
the uevent is like this:
UDEV [1060.112118] add /devices/virtual/vfio/81 (vfio)
ACTION=add
DEVNAME=/dev/vfio/81
DEVPATH=/devices/virtual/vfio/81
MAJOR=509
MINOR=1
SEQNUM=8544
SUBSYSTEM=vfio
USEC_INITIALIZED=1060111894
UDEV [1060.122089] bind /devices/pci0000:d7/0000:d7:00.0/0000:d8:00.0 (pci)
ACTION=bind
DEVPATH=/devices/pci0000:d7/0000:d7:00.0/0000:d8:00.0
DRIVER=vfio-pci
ID_MODEL_FROM_DATABASE=PCIe Data Center SSD (DC P3700 SSD [2.5" SFF])
ID_PCI_CLASS_FROM_DATABASE=Mass storage controller
ID_PCI_INTERFACE_FROM_DATABASE=NVM Express
ID_PCI_SUBCLASS_FROM_DATABASE=Non-Volatile memory controller
ID_VENDOR_FROM_DATABASE=Intel Corporation
MODALIAS=pci:v00008086d00000953sv00008086sd00003703bc01sc08i02
PCI_CLASS=10802
PCI_ID=8086:0953
PCI_SLOT_NAME=0000:d8:00.0
PCI_SUBSYS_ID=8086:3703
SEQNUM=8545
SUBSYSTEM=pci
USEC_INITIALIZED=1060121805
Have tested several kernel versions such as v3.10, v4.10, v4.15, v4.19.
We didn't see an event which is like this:
ACTION=add
DRIVER=vfio-pci
Change-Id: I7299a2fb4d634edaa6bab3412ee8f363f66aae6f
Signed-off-by: JinYu <jin.yu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/452053
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Move actual submission code to _spdk_bdev_io_do_submit, used by
both normal submission path and QoS path.
Previous patch(review.gerrithub.io/c/442127) adds the missing
bdev_io->internal.in_submit_request flag to QoS submission path.
But QoS submission path doesn't handle nomem_io yet.
This patch makes QoS submission path handle nomem_io in the same way
as the normal path and extracts actual submission code into do_submit
function, so that further modification of the submission logic will
apply to both paths automatically.
Change-Id: I41fa88d239c3a2bd9783d812826e32e7c887818d
Signed-off-by: lorneli <lorneli@163.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455252
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: GangCao <gang.cao@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Make the old name a deprecated alias.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ibbf50676e0d989b67121e465fc140f94faec46ed
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453033
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
We flush a cache buffer once it's filled. When the write
for that cache buffer has completed, we look to see if
there's more data to flush. Currently if there's *any*
more data to flush, we will flush it, even if it's not
a full buffer.
That can hurt performance though. Ideally we only want
to flush a partial buffer if there's been an explicit
sync operation that requires that partial buffer to be
flushed. Otherwise we will end up writing the partial
buffer to disk, and then come back and write that data
again later when the buffer is full.
Add a new unit test to test for this condition. This
patch breaks one of the existing unit tests which was
designed specifically around a RocksDB failure condition.
Change that file_length unit test to now write exactly
one CACHE_BUFFER, which still tests the general logic
making sure that we don't confuse the amount of data
flushed with the value written to the file's length
xattr.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I83795fb45afe854b38648d0e0c1a7928219307a2
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455698
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Data can get implicitly flushed as cache buffers are filled. But
the length xattr is only written in response to a sync or close
operation. So we cannot just look at the amount of data flushed,
and ignore the sync operation if all of the data written has been
flushed - we still need to write the length xattr.
This also adds a unit test which reproduces the original problem.
Fixes issue #297.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Icca6ef4d1544f72e9bc31c4ee77d26b4b7f0cce4
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455692
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Interpret volume->open() options as our vbdev_ocf_base struct.
This is used during metadata probe procedure, when there is no
vbdev configurations created, so we need to pass base structure
as option.
This patch is a prepartion for persistant metadata support,
and it does not change current behavior.
Change-Id: I0b7435df1692d8b3028931c6c9fc50d2d84b2557
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455415
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Hotremoving a bdev exposed as a vhost-scsi LUN which
has an allocated io_channel is broken now. By detaching
LUNs from their SCSI devices at the start of LUN
hotremoval, we can end up with the LUN's descriptor
completely leaked.
Upon hotremove, each LUN fires a hotremove_cb to the
upper layer (iscsi/vhost) and starts a poller to
periodically check if lun->io_channel has been closed
already. While iSCSI allocates and frees io_channels
for particular LUNs separately using
spdk_scsi_lun_allocate_io_channel(), vhost allocates
and frees io_channels for the entire SCSI device using
spdk_scsi_dev_free_io_channels(), as it knows there's
only one LUN per SCSI device.
Since LUNs are removed from SCSI devices at the start
of LUN hotremoval, the spdk_scsi_dev_free_io_channels()
is not able to close the io_channel of the child LUN.
That LUN is no longer referenced there, and the function
returns immediately without any error.
By reverting this patch, we practically defer removing
LUN from SCSI device until the bdev descriptor is closed,
that is, until that spdk_scsi_dev_free_io_channels()
and the subsequent spdk_scsi_dev_destruct() in vhost
is called.
Historically, spdk_scsi_lun_free_io_channel() was
introduced after vhost-scsi already supported hotremove,
so before we can introduce a change like this one we
have to refactor vhost to use the new APIs first.
This reverts commit 497997bc60.
Change-Id: I58a9988a22694f5b5b927173098ac8218270462e
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/455371
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Pawel Kaminski <pawelx.kaminski@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
