When removing a listener, there's a small window when the listener is
already freed, while the controllers that were created on that listener
are still active. It happens, because the listener is removed before
disconnecting its qpairs and in turn destroying the controllers.
The ctrlr->listener pointer is now cleared when a listener is freed and
its value is checked for NULL before each use.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Iace65a46eebc5972cc4ef0f1c1822ffc5d3e559e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9237
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
An error might occur after succesful transport creation
when the new transport is added to nvmf poll groups, e.g.
in nvmf_transport_poll_group_create. In that case
transport is not detroyed and poll groups are not fully
functional. To correct this behaviour, destroy transport if
spdk_nvmf_tgt_add_transport fails. Also update nvmf_tgt
initialization step to check that all poll groups were
created.
Change-Id: I116e6944729d846c1755c2844c77825f65db8c12
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9255
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jacek Kalwas <jacek.kalwas@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This is needed for reporting additional information in JSON RPCs
Change-Id: I45da2ea78cd5415f7536130b6793ca29e929b86e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9338
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
When poll_group is used, several qpairs share the same CQ
and it is possible to receive a completion with error (e.g.
IBV_WC_WR_FLUSH_ERR) for already disconnected qpair
That happens due to qpair is destroyed while there are
submitted but not completed send/receive Work Requests
To avoid such situation, we should not detroy ibv
qpair until we reap completions for all submitted
send/receive work requests. That requires some
rework in rdma transport and will be implemented
later
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: Idb6213d45c2a7954b9ab280f5eb5e021be00505f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9056
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Poll group holds lists of qpairs in different states and
when we got rdma completion with error, we iterate these
lists to find a qpair which qp_num matches. qp_num
is stored inside of ibv_qp which belongs to spdk_rdma_qp
structure. When nvme_rdma_qpair is disconnected, pointer
to spdk_rdma_qp is cleaned but qpair may still exist in
poll group list and when we start searhing for qpair by
qp_num we may dereference NULL pointer.
This patch adds a check that pointer to spdk_rdma_qp
is valid before dereferencing it. To minimize boilerplate code,
wrap all check in macro. Add unit test to verify this fix.
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: I1925f93efb633fd5c176323d3bbd3641a1a632a9
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9050
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
When a subsystem is being deleted, we disconnect all qpairs
and when the last qpair for some controller is disconnected,
we start controller desctruction process. This process requires
to send a message to subsystem's thread to remove the controller
from the list in the subsystem and after that send a message to
controller's thread to release resources.
The problem is that the subsystem also destroys all attached
controllers. This order is unpredictable and we may get
heap-use-after-free or double free.
To fix this problem we can rely on the fact that the subsystem
can only be destroyed in incative state, that means that all
qpairs linked to the subsystem are already disconnected and
all controllers are already destroyed or in the process of
destruction.
spdk_nvmf_subsystem_destroy API is now can be asyncrhonous,
it accepts a callback with cb argument.
Change-Id: Ic72d69200bc8302dae2f8cd8ca44bc640c6a8116
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6660
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
John Kariuki tested this patch on a system with
several Intel P5800X Optane SSDs, to determine the
performance impact of adding these two
spdk_trace_records() in the main NVMe I/O path.
The pathological case (512B random reads on a single
Xeon core) decreased from 13.10M to 12.88M, or 1.7%.
Normal workloads (4KB+) would incur a smaller penalty
since the I/O rate would be much lower - maybe even
unnoticeable..
This is a really valuable tracepoint to have enabled
by default, so I think this small amount of degradation
is acceptable.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie2543cadf3541eb74398d31ac0f495522ab49ec0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9303
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
The issue is that:
We call cb_fn firstly, then add the related idxd_ops into the
chan->ops_list. Then if in user's callback, we call spdk_idxd_submit_*
function gain, then we will allocate new idxd_opts first. So if there is
recursive spdk_idxd_submit_* in the call back function, then
we will exhaust all the ops_pool resources. And the function will
report -EBUSY issue in _idxd_prep_command function.
And this patch fixes this issue by adding back idxd_ops before calling
user's call_cb function.
Change-Id: I32239fb29be875bc452803a48dc24ff5fa533016
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9401
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Add more information to users which implementation cannot
be set for idxd.
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: Ib26f117dc19fa1f28cc1d548549caba0b696a915
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9399
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Paul Luse <paul.e.luse@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>
Similar is already done for json-rpc bdev_get_bdevs, it might be
useful for the upper layer which has no interest in all but only
in one specified.
Signed-off-by: Jacek Kalwas <jacek.kalwas@intel.com>
Change-Id: I25f7d20ff953e612d7982207d57607f1c3bbaa90
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9361
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
This new API signals that the ctrlr will soon be
reset. This allows the transport to skip unnecessary
steps in following calls to the driver prior to the
reset - for example, skipping PCIe DELETE_SQ/CQ
commands when freeing an IO qpair.
Note that if we are deleting a qpair after
prepare_for_reset was called, and the qpair is
still waiting for a CREATE_IO_CQ or CREATE_IO_SQ,
we cannot poll for those commands to complete,
but we also cannot free the qpair immediately.
So set a flag for this case to defer the
destruction until the outstanding CREATE_IO_CQ or
CREATE_IO_SQ callback is invoked (typically as an
aborted command when the reset happens).
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I34c6276ae71e7d61ad4a3720f1a985b1ee96bd8b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9249
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Rename endpoint->fd to ->devmem_fd to better reflect its purpose.
Signed-off-by: John Levon <john.levon@nutanix.com>
Change-Id: Ic73cb6684228d4ecb3d58c1865c80a3d7d07c223
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9402
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This patch moves schedueler and governor related API from
the internal event.h to public scheduler.h.
With this it is possible to create subsystem responsible
for handling the schedulers.
Three schedulers and a governor were moved to scheduler modules
from event framework.
This will allow next patch to add JSON RPC configuration
to the whole subsystem.
Along with easier addition of other schedulers.
Removed debug logs from gscheduler, as they serve little purpose.
Signed-off-by: Krzysztof Karas <krzysztof.karas@intel.com>
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I98ca1ea4fb281beb71941656444267842a8875b7
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6995
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
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>
Make sure to cancel the scheduling during:
1) gather_metrics error (already present)
2) setting the scheduler to NULL (new)
3) shutdown of the application (new)
In all of the above scheduler cannot proceed to
the balance() function.
Prior to this patch lack of setting
g_scheduling_in_progress to false would block
_start_subsystem_fini() from proceeding.
Resuling in application never shutting down.
thread_infos are allocated during gather_metrics,
and freed on threads_reschedule.
The added cleanup is used during 1) and 2).
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I5ab77c85aaa9f94bd4e1a1660b55ab2f2c47bbdc
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9326
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
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 report other errors in spdk_interrupt_register(), so add one if there is an
issue with adding the given efd to the fd group.
Signed-off-by: John Levon <john.levon@nutanix.com>
Change-Id: Idcd810d2b3f644bab1514063b399b9767797130f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9388
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
transport specific options are already introduced however dump was missed
Signed-off-by: Jacek Kalwas <jacek.kalwas@intel.com>
Change-Id: I2fe533898d82534cc6cb8a3e5de4e2b2c72ae00a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9355
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Expose the already existing nvme_ctrlr_get_cc as
spdk_nvme_ctrlr_get_regs_cc, similar to spdk_nvme_ctrlr_get_regs_csts and
spdk_nvme_strlr_get_regs_cap etc.
Signed-off-by: Tomasz Bielecki <tomasz.bielecki@wdc.com>
Change-Id: Ibfcf6fbe64dee3719f381184fb728ab6e4d52526
Signed-off-by: Tomasz Bielecki <tomasz.bielecki@wdc.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9220
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Krishna Kanth Reddy <krish.reddy@samsung.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Refine ANA state from per subsystem listener to per subsystem listener
per ANA group.
Add an array of ANA state per ANA group to subsystem listener. The array is
indexed by ANA group ID - 1.
Then in I/O paths, we get ANA state by
ctrlr->listener->ana_state[ns->anagrpid - 1].
The NVMe specification indicates the existence of NVM subsystem specific
ANA state when FFFFFFFFh is specified as NSID for the Get Features
and the Set Features commands. For these, we return the optimized state.
Update the nvmf_subsystem_get_listeners RPC to return all ANA states
of the underlying ANA groups. The nvmf_subsystem_get_listeners RPC is
not matured and not used in the test code yet. Hence compatibility is
not high priority.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ia2d4d5361ac01236f595c22765fd35e4c5fdee0e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9064
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This is the first patch in the patch series to control ANA states not only
as a unit of subsystem listener but also as a unit of ANA group and create
user preferred mapping between namespaces and ANA groups within a single
subsystem.
This patch adds anagrpid to both spdk_nvmf_ns and spdk_nvmf_ns_opts, and adds
ana_group array to spdk_nvmf_subsystem to count number of namespaces per ANA
group within a single subsystem. The size of the ana_group array is equal
with the size of the namespaces.
For each subsystem, allocate ana_group array regardless of the value of
ana_reporting of the subsystem.
For each namespace, at its creation, initialize anagrpid explicitly to be equal
with nsid by default and increments the corresponding entry of the ana_group
array of the subsystem regardless of teh value of the ana_reporting of thee
subsystem.
Hence the contents of the created ANA log page is not changed even if the
algorithm to crete ANA log page is changed.
Additionally this patch adds a unit test case that one ANA group
has multiple namespaces.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I78539db4e7248c2953c6927ff8128cb5a7e34b96
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9102
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
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>
The next patch will add anagrpid to spdk_nvmf_ns_opts. This patch
is for the upcoming change and futhre potential changesto ensure the
ABI compatibility.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ibe1024a27ea4ab823d40d775470cde78b9eb8d2c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9112
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Previously we used assert for this check.
Fix#2141.
Change-Id: Ieaf8e34721a5b65e6d7790d56587239e2e6d75f4
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9383
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
If a response is returned prior to _nvmf_request_complete being called then the cid in the response is
not set correctly and the PDU state is not reset which causes a hang and the PDU state machine is
expecting more data but none will be sent. There are two cases where this can occur:
1) If the request is bi-directional
2) If nvmf_tcp_req_parse_sgl returns an error (e.g max_io_size exceeded)
Signed-off-by: matthewb <matthew.burbridge@hpe.com>
Change-Id: Icc3ed02a4499a12d8920e6433a746b72022a72fe
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9327
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
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>
If NN is very, very large, this allocates too much memory. For now, just
use a list.
Change-Id: I904977673d8fb6c86f03c94ba798c6cc07f4a4d8
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9301
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
This can be done immediately after receiving the controller identify
data for now.
Change-Id: I527a44c4d1f4d3ad2eeb8fc77e07086c2358cac3
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9300
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Interrupts were all being registered with the name "fn". Directly pass in the
name instead to fix this.
Signed-off-by: John Levon <john.levon@nutanix.com>
Change-Id: I38e59d784298a61ca2130090ede500101389685c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9346
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-by: Rui Chang <rui.chang@arm.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
The root cause is that we do not the copy the contents
to the dst address correctly.
So provide a _sw_accel_copyv function to address this
issue.
Fixes#2096
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: I25d0e60f51abd41ed77b2b23f08d609f4f052de0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9193
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Monica Kenguva <monica.kenguva@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Previously the accel_perf tool would look at whether it had HW
or SW commands to know whether to execute the callback right away
or schdule to avoid blowing the stack (SW calls are sync).
Moved this to the SW module (part of the accel engine) so the
caller doesn't have to worry about. Allowed for a few simplifcations
in the tool as well.
Also, instead of using send_msg to call the completion we add it to
a list in the sw module that a new poller uses to perform the
completions as this is more efficient the sending a message to the
same thread.
Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: Ifc6c5b8635f51e3fa1a825c8573378b3752d7d91
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9171
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
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: Jim Harris <james.r.harris@intel.com>
Small but noticable on perf top. We were using a list to track
valid batches and checking it on every submission. Instead just
put a valid channel ptr in the batch struct and clear it when the
batch is freed. There was no other reason for the list.
Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: Ia26ceee578be4a82f4fd8abb2358ff18c56271a4
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9149
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
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>
There is no need to track if governor was enabled during
initialization of the scheduler dynamic.
Instead _spdk_governor_get() can be used to determine that.
While here set the default g_governor to NULL.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Ibdfde3d2714dd0b6d629f19f0d246b4fc3c9d214
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9325
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This patch cleans up the header file, structures and
parameters of governor API. While documenting the
functionality.
- made governor name const
- renamed _spdk_governor_list_add() to _spdk_governor_register()
This is preparation to making this API public.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Ie394109c839dead0e7ade946f95be8105b00e674
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8843
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This patch cleans up the header file, structures and
parameters of scheduler API. While documenting the
functionality.
- made scheduler name const
- removed typedefs for schedueler callbacks
- balance() now accepts uint32_t for array size instead of an int
- removed unused _spdk_lw_thread_set_core()
- renamed _spdk_scheduler_period_set() to _spdk_scheduler_set_period()
- renamed _spdk_scheduler_period_get() to _spdk_scheduler_get_period()
- renamed _spdk_scheduler_list_add() to _spdk_scheduler_register()
This is preparation to making this API public.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Ia7b6b6a5eafb052ac275db6c04113a8ad442383f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8842
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Registering multiple governors would fail due to them having
the same name. Only saved by the fact that right now,
there is only one governor registered in this fashion.
Fix it by adding name of the governor structure passed
to the function name.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Ic7a206da2c8f5dc1e72e41629bccf989c030f182
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8792
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@gmail.com>
By default g_scheduler is now NULL. It can be set
either by event framework or RPC.
To keep RPC consistent, the 'static' scheduler was kept.
All access to the g_scheduler is done via _spdk_scheduler_get().
Access to its members is done to balance, get name for RPC and
through _spdk_scheduler_set(). All of them happen on scheduling
reactor and don't pose race condition when changing scheduler.
There is no need to delay init/deinit of scheduler via g_new_scheduler.
This variable was removed and all that happens in _spdk_scheduler_set().
To unset and deinitialize current scheduler,
_spdk_scheduler_set(NULL) has to be called.
This results in moving scheduler deinitalization to that
call too.
Every spdk_scheduler callback is now mandatory.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I1fed766989de9bcaeb7e82b490c1275f3572d77d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8811
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@gmail.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
By default g_governor is now NULL. It can be set
either by event framework or schedulers directly.
Dynamic_scheduler and gscheduler specifically want
to use the dpdk_governor, so their initialization
now sets it explicitly.
To unset and deinitialize current governor,
_spdk_governor_set(NULL) has to be called.
This results in moving governor deinitalization to that
call too.
The "default" governor has been removed.
Every spdk_governor callback is now mandatory.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Ibf76bd28bfbb159416026996fa217bb3325a3d31
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8810
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@gmail.com>
There is no explicit need for the spdk governors initialization
to occur on per core basis.
This implementation detail for dpdk_governor is now hidden
in the init/deinit calls. There is no recourse when failing
deinit for a certain core, so ignore the return code.
Changed return type for deinit in governor and scheduler to void.
While here modified the callbacks for scheduler to no
longer require passing currently selected governor as an argument.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I7f0b7a09aa7f5d12ae47fca25186faeedac31a95
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8791
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@gmail.com>
Add the new device ID for VMD devices so VMD devices
can be unbound and used with the SPDK setup script.
Bus numbering for VMD devices is different on IceLake platforms,
and only half of the bus numbers are available. Add a function to
set the starting bus number and the max bus number by reading the
new BUS_RESTRICT_CAP and BUS_RESTRICTIONS VMD registers.
Signed-off-by: Sydney Vanda <sydney.m.vanda@intel.com>
Change-Id: I8905d4bcba84c74e3dadfb27262e668c4281b0c8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8331
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
When deleting an IO qpair, make sure that it's connection process is
finished (i.e. create CQ/SQ commands are completed) before freeing it.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I487dcef390d73ff4a7264ff97d965c9030916840
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9279
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This allows for creating admin qpair in an asynchronous mode and I/O
qpairs based on what the user specified in spdk_nvme_io_qpair_opts.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I1801ea76d5a08b1bdc558eb0f18648f59454b654
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9077
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Monica Kenguva <monica.kenguva@intel.com>
Split the connection process across two states, which allows the
transport to connect the admin queue asynchronously.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Ie84477331df0abf5ffdfc2a0ff5d5ada760c9e73
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/9076
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Monica Kenguva <monica.kenguva@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This allows for submitting IO requests before the CONNECT command is
sent and not stopping the connect process due to the CONNECT being
queued. It could happen once IO qpair connection is asynchronous.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I04e45b1c2f49f9da3412c843ea899341c56a0420
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8624
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
