Commit Graph

7739 Commits

Author SHA1 Message Date
Seth Howell
33668b2254 rdma: change structure of drained_qpair to work w/ messages.
This will become important later on.

Change-Id: I94e5af03359e476afbc68664e43f44269ad5974c
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448074
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-18 23:32:21 +00:00
Seth Howell
7dd3cf441a rdma: limit the completion queue based on the SRQ.
When we have a shared receive queue, the number of outstanding items
associated with a completion queue is deterministic, and limited by how
many RECVs we have total in the SRQ. So, we can set the total size of
the Completion queue at the beginning of time and never resize it.

Change-Id: I787e4c5bbd52ac8948a323d1301f926f887cd91c
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447492
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>
2019-03-18 23:32:21 +00:00
Seth Howell
a5972c6245 rdma: consolidate common error paths in qpair_init
Consolidating error paths is common practice in SPDK so do that here to
make the function more uniform and save space.

Change-Id: I98c5d5f7feeb688f1d8b24f4d2d3461a43d00c1d
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448191
Reviewed-by: Jim Harris <james.r.harris@intel.com>
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>
2019-03-18 23:32:21 +00:00
Seth Howell
97a43680a9 rdma: move cq_resize to its own function.
Change-Id: I07aef399320fd4a014f63760670ea765d2e18b4b
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448190
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>
2019-03-18 23:32:21 +00:00
Seth Howell
fa79f64ad1 rdma: Keep a pointer to the SRQ in the qpair
Change-Id: Id173038b6ad6b1564acf5d6886814f7d310964c7
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447471
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>
2019-03-18 23:32:21 +00:00
Seth Howell
01201d3e87 rdma: remove compile time config for SRQ
Change-Id: I44af3ee4dc6ec76045e1d0614910402487098a3d
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447120
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>
2019-03-18 23:32:21 +00:00
paul luse
cf467578f8 bdev/reduce: fix issue with vol_load
Discovered in bdev compression test.  Vol load was using the index
of the logical map to set the allocated chunk map bits as well as
retrieve the chunk. Exposed with overwrite situations. Also clarrified
one var name that I missed in previous patch.

Also added UT that fails without the fix, passes with the fix.

Change-Id: I776a5af9364cc3d4e8b7dafec22acf23bd463630
Signed-off-by: paul luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447969
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>
2019-03-18 17:41:32 +00:00
paul luse
228396f898 lib/reduce: change some var names
Changed a few cases where var name chunk was used where a better
name would be logical_map_index to avoid confusion with 'chunk'
used as a data pointer.

Change-Id: Ib82a783725cad4609c31bfa8997debb3e71e75b9
Signed-off-by: paul luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447962
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>
2019-03-18 17:41:32 +00:00
Darek Stojaczyk
07042bdeb1 env/dpdk: remove the fake rte_bus
This practically reverts commit 2fe7aa5e4 [1].

The extra rte_bus was supposed to allow running SPDK
as a non-priviledged user by enabling RTE_IOVA_VA mode.
DPDK uses RTE_IOVA_PA by default - which means there are
physical addresses used as memseg IOVAs and hence the
root access is required to retrieve those physical
addresses. This patch was supposed to be paired with
a different DPDK patch of mine, but DPDK rejected that
one. Instead, in DPDK 18.11+ the user can force iova mode
by specifying --iova-mode=<mode> command line option,
where <mode> is either pa or va.

Either way, apparently there are cases where physical
address contiguity is required even without UIO (#707)
so let's revert this patch and consistently stick with
RTE_IOVA_PA.

SPDK requires some more effort to support running as
a non-priviledged user anyway.

Fixes #707

[1] env_dpdk/vtophys: register a fake rte_bus supporting RTE_IOVA_VA

Change-Id: I90bf097fd6a7f129444229dc7cf07a462d4f2d09
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448121
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>
2019-03-18 16:36:40 +00:00
Pawel Niedzwiecki
ac2dd4b2c2 test/ftl: Change fio_plugin from basic test to bdevperf
Ftl tests won't pass with fio_plugin when asan is enabled.

Change-Id: I6f07f661c19ecab302e291bbd76a7aad964000c7
Signed-off-by: Pawel Niedzwiecki <pawelx.niedzwiecki@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447318
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2019-03-18 16:33:49 +00:00
Wojciech Malikowski
11c8fbcf62 lib/ftl: Fix memory leak in restore module
Change-Id: I39c89ef935eeac56fd860b11e1fafd5047072f7e
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448023
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-18 16:33:49 +00:00
Wojciech Malikowski
d058328e12 test/unit/ftl: Add unit tests for metadata serialization
Change-Id: I1f96efc289d4804677b081a1cc6866983ca8eb77
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/445875
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
2019-03-18 14:04:29 +00:00
yidong0635
d47468a151 test scripts: add rpm-build package in vm_setup.sh
This package is needed in vm_setup.sh for vpp,
without this will report "rpmbuild: command not found".

Signed-off-by: yidong0635 <dongx.yi@intel.com>
Change-Id: Ic00a55addc940e359aae17b45e58b9a736b08fd3
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447282
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Piotr Pelpliński <piotr.pelplinski@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
2019-03-18 12:55:02 +00:00
Darek Stojaczyk
c2a53da73c vhost: don't check indirect desc feature flag
Windows Virtio drivers use indirect descriptors without
negotiating their feature flag, which is explicitly
forbidden by the Virtio 1.0 spec. "(2.4.5.3.1 Driver
Requirements: Indirect Descriptors) The driver MUST NOT
set the VIRTQ_DESC_F_INDIRECT flag unless the
VIRTIO_F_INDIRECT_DESC feature was negotiated.".

Violating this rule doesn't cause any issues for SPDK
vhost, but triggers an assert, so we can only run Windows
VMs with non-debug SPDK builds.

This patch removes the assert and allows Windows VMs
to be run with debug versions of SPDK vhost.

Fixes #650

Change-Id: I95f534c33c384a4e1126a8c343c21eb63ec7bcef
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447803
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2019-03-18 05:54:16 +00:00
Jim Harris
5a77afa8fb test/bdev/raid: reduce memory usage for io_waitq unit test
This test allocates ridiculous amounts of memory for certain
random seeds.  This test doesn't benefit from the extra queue
depth, or the larger I/O sizes, so just pick more reasonable
numbers that suffice for this unit test.

Some of the failing seeds in the test pool resulted in >30GB
of total memory allocated, according to valgrind.

Eventually we need to remove the randomness from all of the unit
tests, but at least this gets us past this intermittent failure.

Fixes issue #692.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia882fda8ef449c8537f76c878008aaf469ae7bd7

Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448193
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2019-03-18 05:53:12 +00:00
Pawel Wodkowski
05085f6645 vhost-scsi: don't return SCSI device if it under removal
spdk_vhost_scsi_dev_get_tgt() shouldn't return devices that are under
removal process.

Change-Id: I22a3f5e811c90244e9640d8d9c65cd7a65a6ecbf
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447071
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2019-03-15 22:06:42 +00:00
Pawel Wodkowski
b1a6197068 vhost-scsi: change removed flag into small state machine
Introduce EMPTY, PRESENT and REMOVED states for Vhost SCSI targets. This
does not introduce any functional changes but opens a way to stop using
both removed flag and spdk_scsi_dev pointer as indicator if device is
removed or not.

Change-Id: Iecd76ffe9e8121cc1359b1e268eb21679d13598e
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447070
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2019-03-15 22:06:42 +00:00
Darek Stojaczyk
b9b1c9592e vhost/compat: implement SET/GET_CONFIG
rte_vhost has rejected a patch with this feature, so
we implement it using the external rte_vhost msg handling
hooks directly in SPDK.

Change-Id: Ib072fc19b921fe0fa01c7f4892e60430232e3a1c
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447025
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 22:03:26 +00:00
Darek Stojaczyk
aa5f129f57 vhost: don't operate on partially-initialized vdev
Make the vdev initialization happen before calling
any vdev related functions. This is mostly needed
for an upcomming patch where additional step is
required after initializing the vdev and before
starting rte vhost.

On the other hand, this patch also fixes a technically
possible scenario where rte vhost starts processing
vhost-user messages and calling our ops before the
related vdev was initialized.

Change-Id: I8fbc7e7bc0b364327cfcec60faa74d4f64d6fad8
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447024
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 22:03:26 +00:00
Darek Stojaczyk
1ecbb6a8da vhost/compat: start polling queues prematurely
rte_vhost requires all queues to be fully initialized
in order to start I/O processing. This behavior is not
compliant with the vhost-user specification and doesn't
work with QEMU 2.12+, which will only initialize 1 I/O
queue for the SeaBIOS boot. Theoretically, we should
start polling each virtqueue individually after
receiving its SET_VRING_KICK message, but rte_vhost is
not designed to poll individual queues. So we use
a workaround to detect when a vhost session could be
potentially at that SeaBIOS stage and we mark it to
start polling as soon as its first virtqueue gets
initialized. This doesn't hurt any non-QEMU vhost slaves
and allows QEMU 2.12+ to boot correctly. SET_FEATURES
could be sent at any time, but QEMU will send it at
least once on SeaBIOS initialization - whenever
powered-up or rebooted.

Vhost sessions are still mostly started/stopped from
within rte_vhost callbacks, but now there's additional
concept of "forced" polling, in which SPDK starts
sessions manually, while rte_vhost still thinks the
sessions are stopped. This can potentially lead to cases
where a session is "started" twice, or gets destroyed
while it's still being polled (by force). Those cases
also need to be handled within this patch.

Change-Id: I70636d63e27914906ddece59cec34f1dd37ec5cd
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446086
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 22:03:26 +00:00
Changpeng Liu
d11aa87320 nvmf: add reservation information to each subsystem's poll group
Change-Id: Idcbc3053daf756c818ae3715b4ba0cbd91ed3d44
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446212
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>
2019-03-15 20:45:43 +00:00
Changpeng Liu
2099401e94 nvmf: rename subsystem poll group's num_channels to num_ns
Array channels in the subsystem's poll group are indexed by
nsid - 1, so rename the previous num_channels to num_ms
makes more sense.  Also embed the channels into a namespace
data structure here, and this can be reused in the following
patch.

Change-Id: If5d9aab4b1d5bcf7a3c22f29fa58d84752f0d4cc
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446211
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>
2019-03-15 20:45:43 +00:00
Ziye Yang
237260f6a1 nvme/tcp: Restore nvme_tcp_read_data function.
Purpose:  To fix performance degradation.
Restore the nvme_tcp_read_data function, and for
nvme_tcp_readv_data, will call nvme_tcp_read_data
if iovcnt=1.

Change-Id: Ie3abb943c294dc27d736371cd7ce9a5aa966877a
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448030
Reviewed-by: wuzhouhui <wuzhouhui@kingsoft.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2019-03-15 19:34:58 +00:00
Darek Stojaczyk
ec2cb43589 barrier: cleanup the file
Group all definitions for a specific architecture
under a single #ifdef. This makes the code more
readable.

Change-Id: I4e88b9bb63f84b6275f10731aa657739ce6862e3
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447881
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 19:32:55 +00:00
Seth Howell
0d3fcd10e9 rdma: add function to create qpair resources.
Change-Id: Id865e2a2821fe04c1f927038d6dd967848cd9a55
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446999
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 19:19:17 +00:00
Ben Walker
353fbcdaf0 nvmf/rdma: Create function to destroy rdma resources
This unifies the clean up path between SRQ and normal
operation.

Change-Id: I396d7e3749579f27b5bb1e89b9d6761a77ba5beb
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446979
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2019-03-15 19:19:17 +00:00
Ben Walker
b25751d99d nvmf/rdma: Add a structure to hold rqpair/rpoller resources
Depending on whether SRQ is enabled, resources may be allocated
to the rqpair or to the rpoller. Create a struct to hold these
pointers that can be used in both locations to avoid duplicated
code.

Change-Id: I2c8fc59009201d9e41721e6462a81732b529a9e0
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446978
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Eugene Kochetov <evgeniik@mellanox.com>
2019-03-15 19:19:17 +00:00
Ben Walker
527be2bf4e nvmf: Remove qpair_is_idle
This wasn't used anywhere.

Change-Id: I405af3c808be284d19218f3f04c1e90e33e31de8
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446977
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
2019-03-15 19:19:17 +00:00
Evgeniy Kochetov
ed0b611fc5 nvmf/rdma: Add shared receive queue support
This is a new feature for NVMEoF RDMA target, that is intended to save
resource allocation (by sharing them) and utilize the
locality (completions and memory) to get the best performance with
Shared Receive Queues (SRQs). We'll create a SRQ per core (poll
group), per device and associate each created QP/CQ with an
appropriate SRQ.

Our testing environment has 2 hosts.
Host 1:
  CPU: Intel(R) Xeon(R) CPU E5-2609 0 @ 2.40GHz dual socket (8 cores total)
  Network: ConnectX-5, ConnectX-5 VPI , 100GbE, single-port QSFP28, PCIe3.0 x16
  Disk: Intel Optane SSD 900P Series
  OS: Fedora 27 x86_64
Host 2:
  CPU: Intel(R) Xeon(R) CPU E5-2630 v2 @ 2.60GHz dual-socket (24 cores total)
  Network: ConnectX-4 VPI , 100GbE, dual-port QSFP28
  Disk: Intel Optane SSD 900P Series
  OS : CentOS 7.5.1804 x86_64
Hosts are connected via Spectrum switch.
Host 1 is running SPDK NVMeoF target.
Host 2 is used as initiator running fio with SPDK plugin.

Configuration:
- SPDK NVMeoF target: cpu mask 0x0F (4 cores), max queue depth 128,
  max SRQ depth 1024, max QPs per controller 1024
- Single NVMf subsystem with single namespace backed by physical SSD disk
- fio with SPDK plugin: randread pattern, 1-256 jobs, block size 4k,
  IO depth 16, cpu_mask 0xFFF0, IO rate 10k, rate process “poisson”

Here is a full fio command line:
fio  --name=Job --stats=1 --group_reporting=1 --idle-prof=percpu \
--loops=1 --numjobs=1 --thread=1 --time_based=1 --runtime=30s \
--ramp_time=5s --bs=4k --size=4G --iodepth=16 --readwrite=randread \
--rwmixread=75 --randrepeat=1 --ioengine=spdk --direct=1 \
--gtod_reduce=0 --cpumask=0xFFF0 --rate_iops=10k \
--rate_process=poisson \
--filename='trtype=RDMA adrfam=IPv4 traddr=1.1.79.1 trsvcid=4420 ns=1'

SPDK allocates the following entities for every work request in
receive queue (shared or not): reqs (1024 bytes), recvs (96 bytes),
cmds (64 bytes), cpls (16 bytes), in_capsule_buffer. All except the
last one are fixed size. In capsule data size is configured to 4096.
Memory consumption calculation (target):
- Multiple SRQ: core_num * ib_devs_num * SRQ_depth * (1200 +
  in_capsule_data_size)
- Multiple RQ: queue_num * RQ_depth * (1200 + in_capsule_data_size)
We ignore admin queues in calculations for simplicity.

Cases:
1. Multiple SRQ with 1024 entries:
   - Mem = 4 * 1 * 1024 * (1200 + 4096) = 20.7 MiB
     (Constant number – does not depend on initiators number)
2. RQ with 128 entries for 64 initiators:
   - Mem = 64 * 128 * (1200 + 4096) = 41.4 MiB

Results:
FIO_JOBS   kIOPS     Bandwidth,MiB/s  AvgLatency,us  MaxResidentSize,kiB
       RQ       SRQ     RQ      SRQ    RQ       SRQ      RQ       SRQ
1      8.623    8.623   33.7    33.7   13.89    14.03    144376   155624
2      17.3     17.3    67.4    67.4   14.03    14.1     145776   155700
4      34.5     34.5    135     135    14.15    14.23    146540   156184
8      69.1     69.1    270     270    14.64    14.49    148116   156960
16     138      138     540     540    14.84    15.38    151216   158668
32     276      276     1079    1079   16.5     16.61    157560   161936
64     513      502     2005    1960   1673     1612     170408   168440
128    535      526     2092    2054   3329     3344     195796   181524
256    571      571     2232    2233   6854     6873     246484   207856

We can see the benefit in memory consumption.

Change-Id: I40c70f6ccbad7754918bcc6cb397e955b09d1033
Signed-off-by: Evgeniy Kochetov <evgeniik@mellanox.com>
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/428458
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>
2019-03-15 19:19:17 +00:00
Vitaliy Mysak
f186434ec2 ocf/test: update integrity test for multicore case
Change integrity test with FIO
  to use multicore configuration.

Change-Id: I0527d9e7832c1a686b9b4fb20cb317ed17acb051
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/445098
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2019-03-15 19:04:10 +00:00
Vitaliy Mysak
00565ee6b5 ocf/test: add management tests for multicore cases
Add functional tests on creation and deletion of
  of OCF bdevs with many-to-one configuration.

Change-Id: I6b740605a3341cb95e8d2a4173eeac04b0ae832b
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/445097
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2019-03-15 19:04:10 +00:00
Vitaliy Mysak
97d1311fb0 ocf: multicore support
This patch adds support for many-to-one configuration
  for OCF bdev. Meaning that one cache instance can
  be used to cache multiple core devices.

Without this change, it is not possible
  to attach core device to existing cache instance.

Change-Id: Idfce023c08b6d41490d590a8abb51f626252cd4d
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/444277
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2019-03-15 19:04:10 +00:00
Wojciech Malikowski
5d82fa6af7 lib/ftl: Pass errors from nvme layer only in completion callback
FTL should pass IO errors to upper layer only in
completion callback to be consistent on sync and
async error paths.

Change-Id: I29c81af477b476c1e10b7933126737a222553ffa
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447176
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 05:03:39 +00:00
Wojciech Malikowski
3816651f4b lib/ftl: Try to process more then one I/O completion
This patch allows to process more then one I/O
completion per reactor iteration.

Change-Id: Ib58013cd96c2b7fe378f175b34d58c6d1d219dcd
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447175
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Young Tack Jin <youngtack.jin@circuitblvd.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2019-03-15 05:03:39 +00:00
Pawel Wodkowski
5c6ca5b6e0 jsonrpc: add Python with statement support
The JSONRPCClient class now support the with statement. From now on the
__del__() method is changed to close() and user need to call it manually
at object disposal or use the 'with' Python statement.

Change-Id: I74afd93d411b9596ab8f9523c54a4a7523eec5e5
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/444686
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Pawel Kaminski <pawelx.kaminski@intel.com>
2019-03-15 05:01:27 +00:00
Pawel Wodkowski
a282588584 scripts/rpc: support sending multiple requests
This will allow testing notifications code where user will most likely
issue many requests and socket buffereing might just squash them into
one packet.

Note that multiple requests are not send in batch mode.

Change-Id: Icfa4bfe21ff1268796bc41cf0443737f260b6040
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/436529
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 05:01:27 +00:00
Pawel Wodkowski
27cc63ec68 rpc: extend logging in JSONRPCClient class
Add extra long argument '--verbose <VERBOSE LEVEL>' to add more log
levels Default is error. Legacy'-v' parameter is translated into INFO
log level.

Some parts of code are logged using DEBUG level.

Change-Id: I2aaa6e2fbaf9b2101c6eeaec0cef6141636f135b
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/436528
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-15 05:01:27 +00:00
Pawel Wodkowski
d5ac3eb190 scripts/rpc: add support for receiving multiple responses
Switch to raw_decode in recv() function. This allow to decode JSON
objects one by one. Decoded part is removed from from receive bufer.

Change-Id: Id0d78a2ace85bcbb9cc8e30d72da6c1c2cad753c
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/435507
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>
Reviewed-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-by: Pawel Kaminski <pawelx.kaminski@intel.com>
2019-03-15 05:01:27 +00:00
Pawel Wodkowski
336faf1074 scripts/rpc: split JSONRPCClient.call into send and recv
Sending multiple request is doable now. Note that, stream of responses
is not supported yet.

Change-Id: I9c49dbf033c67074e603f2e48a469142c3c63653
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/435481
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>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-by: Pawel Kaminski <pawelx.kaminski@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
2019-03-15 05:01:27 +00:00
Shuhei Matsumoto
8353d56e77 iscsi: Fix performance degradation due to sock_readv
spdk_iscsi_conn_read_data was changed to use sock_readv
instead of sock_recv as a refactoring effort.
However this change caused about 3% performance degradation
for write I/O.

Hence this patch changes spdk_iscsi_conn_read_data to use
sock_recv and changes spdk_iscsi_conn_readv_data to call
spdk_iscsi_conn_read_data if iovcnt is 1.

Change-Id: I5fcad03ff23dee4e9954eee879225d3c493be52e
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448021
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-14 23:09:56 +00:00
Shuhei Matsumoto
1bec7d572c iscsi: Fix the bug when querying DIF context for Data In PDU
spdk_iscsi_get_dif_ctx() didn't work correctly for Data In PDU
because we had to get iSCSI task from PDU directly for Data In PDU,
and lun_id was not copied in spdk_iscsi_task_get().

This patch fixes these bugs and the DIF strip feature was verified
after applying this patch.

Change-Id: I74d404b82c4a9502a9a8e166748f817d3c2e4368
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447884
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@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>
2019-03-14 22:51:11 +00:00
Changpeng Liu
1151e65dc1 nvme: use asynchronous probe API in spdk_nvme_connect()
Change-Id: Ib7b21dc64c2d5a5594b05b4d8ade839fc4b06fbc
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447023
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-14 22:37:02 +00:00
Changpeng Liu
84245b7202 nvme: add spdk_nvme_connect_async() API
Users may want to connect specified controller at running time,
so this API will connect to the controller and return probe context
to users, users must call spdk_nvme_probe_poll_async() to initialize
the controller to the READY state before using it.

Change-Id: I232886b000454ee826ea73c4e1043d0d18ee0ec6
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/445657
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-14 22:37:02 +00:00
Changpeng Liu
8e4d6b3d0d fio_plugin: remove metadata location limitation when doing PI
Change-Id: Ibdf833c6f6de30efa1599d35b4977149441b368a
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446824
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>
2019-03-14 19:30:53 +00:00
Ben Walker
5ffeed2582 configure: Make indentation consistenly use tabs
We don't have enforcement for bash scripts, so of course
this ended up with mixed tabs and spaces.

Change-Id: Ideba6cdffa0687a40851a5cda6216a13f870a9ba
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447242
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2019-03-13 23:19:37 +00:00
Karol Latecki
a42dfab18e scripts/perf: SPDK NVMeOF benchmark scripts
Automation scripts for reproducing scenarios described in
NVMeOF performance report by Vishal.
https://ci.spdk.io/download/performance-reports/SPDK_nvmeof_perf_report_18.04.pdf

Change-Id: I006f68b67b5f42f2cc1ec82a158fc86e76371ba1
Signed-off-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/441898
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2019-03-13 23:19:10 +00:00
Pawel Kaminski
b05645c196 test/hotplug: Use malloc as second bdev instead of nvme in hotremove tests
In test soft hotremove is used.
For easier testing use only one nvme bdev instead of two.

Change-Id: Ia5543f0ae0773e38b92cc4d2325da2e3d787f674
Signed-off-by: Pawel Kaminski <pawelx.kaminski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446528
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2019-03-13 17:57:14 +00:00
Darek Stojaczyk
b0b3d23270 ocf: do not build ocf_inc and ocf_src concurrently
Those targets use the same directories for temporary
files and at the moment they race with each other. We're
seeing intermittent failures on CI:

```
INSTALL [...]/ocf/env/src/ocf/utils/utils_list.c
rm: missing operand
```

```
INSTALL [...]/ocf/env/src/ocf/utils/utils_device.h
INSTALL [...]/ocf/env/src/ocf/utils/utils_io.c
rm: cannot remove [...]/ocf/env/src/ocf/utils': Directory not empty
```

Fixes #709

Change-Id: I9bf6ee7a97144afafbb643a0e350f38abe79270f
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447750
Reviewed-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2019-03-13 17:56:07 +00:00
Changpeng Liu
92ff5e0b12 fio_plugin: export apptag and apptag mask to users
Existing fio plugin tool uses hardcoded application
tag and application tag mask for end-to-end data
protection, we export the two options to users
now.

Change-Id: I64d89c29e99030ce8daa2947e73d941b73ac4a8e
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446384
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>
2019-03-13 17:22:45 +00:00
Changpeng Liu
b5853777fe fio_plugin: enable separate metadata(DIX) support
Existing APIs used in the fio plugin tool already
contain the separate metadata parameter, so we just
need to allocate a separate metadata buffer for each
request, by default, each request will have 4096
metadata buffer size when PI enabled with separate
metadata, but also providing an option here to let
users can input bigger value in case one request
will need larger metadata buffer size.

Change-Id: I51679c5cb7f7b1599b81287b1fbb8d9be7959191
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446375
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>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2019-03-13 17:22:45 +00:00