fuse_dispatcher: make header internal

The fuse_dispatcher is not intended to be a public API, it's for
internal use only.

Signed-off-by: Jim Harris <jim.harris@nvidia.com>
Change-Id: I23e839a8f7155

fuse_dispatcher: make header internal

The fuse_dispatcher is not intended to be a public API, it's for
internal use only.

Signed-off-by: Jim Harris <jim.harris@nvidia.com>
Change-Id: I23e839a8f71557960fe27f83c2eb9e51c57c8ea8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/25516
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Ben Walker <ben@nvidia.com>

show more ...

lib/mlx5: API to configure UMR

Add API to configure regular UMR, without BSF and
any offloads, only scatter-gather functionality

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: Ib4f

lib/mlx5: API to configure UMR

Add API to configure regular UMR, without BSF and
any offloads, only scatter-gather functionality

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: Ib4fb21c5d27c3a89aef649ca6fd0162ba9d10e8a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24706
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Community-CI: Mellanox Build Bot
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

replace strtok with strtok_r

The strtok function is not reentrant. To address this issue in a
multithreaded environment, the strtok_r function should be used,
which is the reentrant version of strto

replace strtok with strtok_r

The strtok function is not reentrant. To address this issue in a
multithreaded environment, the strtok_r function should be used,
which is the reentrant version of strtok.

Signed-off-by: XuQi <1530865185@qq.com>
Change-Id: I35c07c7cf4e20bacb7b1e7c7adaedfcd1a81f86e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/25492
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeliu@tencent.com>
Reviewed-by: Jim Harris <jim.harris@nvidia.com>
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>
Community-CI: Mellanox Build Bot

show more ...

lib/rdma_provider: Add API to check if accel seq supported

verbs provider doesn't support accel sequence
mlx5_dv provider supports accel sequence if a module
which implements UMR is registered, i.e.

lib/rdma_provider: Add API to check if accel seq supported

verbs provider doesn't support accel sequence
mlx5_dv provider supports accel sequence if a module
which implements UMR is registered, i.e. accel_mlx5 driver
is registered

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I59944aceb22661f9de3198ecd571b1a73af28584
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24701
Community-CI: Mellanox Build Bot
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>

show more ...

lib/mlx5: Add API to check if UMR registration supported

The 2 new APIs are used for negotiation between different
modules to use accel sequence/UMR

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.

lib/mlx5: Add API to check if UMR registration supported

The 2 new APIs are used for negotiation between different
modules to use accel sequence/UMR

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: Ifca6a3e3a7649e48392785d7a977e1b468dbf4be
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24700
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>
Community-CI: Mellanox Build Bot

show more ...

lib/rdma_provider: Allow to set data_transfer cb

Extend qpair init attr with data_transfer callback
to be set to the memory domain.
verbs provider doesn't support this functionality
The data_transfe

lib/rdma_provider: Allow to set data_transfer cb

Extend qpair init attr with data_transfer callback
to be set to the memory domain.
verbs provider doesn't support this functionality
The data_transfer callback will be used in next
patches to support accel sequence in nvme_rdma

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: Ibc4aa57057e51ed2e0b687132fdb7a2b4368f5fa
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24693
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Community-CI: Mellanox Build Bot
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>

show more ...

nvme/rdma: Allocate memory domain in rdma provider

Next patches add data_transfer function to a memory domain,
for mlx5_dv provider, that means we can't use a memory
domain created via rdma_utils. I

nvme/rdma: Allocate memory domain in rdma provider

Next patches add data_transfer function to a memory domain,
for mlx5_dv provider, that means we can't use a memory
domain created via rdma_utils. In future, memory domain
will hold a qpair pointer, to minize changes we create
a memory domain per qpair in this patch
The verbs provider still uses rdma_utils library.

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I53c20b70901c1061c8a067c612dc4ce6b9a3999a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24692
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <ben@nvidia.com>
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>

show more ...

event: move struct spdk_lw_thread to internal header

This does not need to be exposed through public API.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: I7f2f1bf5d47e9b0ded011a4f3f83

event: move struct spdk_lw_thread to internal header

This does not need to be exposed through public API.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: I7f2f1bf5d47e9b0ded011a4f3f832320b106b3ec
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/25032
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

lib/event: implement scheduler trace events

Change-Id: I284b54b6a92aa330b5a95700101a928b9d83d46f
Signed-off-by: Anisa Su <anisa.su@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk

lib/event: implement scheduler trace events

Change-Id: I284b54b6a92aa330b5a95700101a928b9d83d46f
Signed-off-by: Anisa Su <anisa.su@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24873
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

include/spdk_internal: define scheduler tracegroup and tracepoints

Change-Id: I50e71c55bcc3b68067b4b5609bef269508ed91d6
Signed-off-by: Anisa Su <anisa.su@samsung.com>
Reviewed-on: https://review.spd

include/spdk_internal: define scheduler tracegroup and tracepoints

Change-Id: I50e71c55bcc3b68067b4b5609bef269508ed91d6
Signed-off-by: Anisa Su <anisa.su@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/25066
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

tcp: make nvme_tcp_qpair_state private

There's no reason to have a common enum for qpair state, as the qpair is
processed differently on the target vs. initiator. The target didn't
even use all of

tcp: make nvme_tcp_qpair_state private

There's no reason to have a common enum for qpair state, as the qpair is
processed differently on the target vs. initiator. The target didn't
even use all of the defined values.

Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Icec6ff56929ca51e7c7c5efd784512d78544631b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24231
Reviewed-by: Krzysztof Karas <krzysztof.karas@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

lib/mlx5: Add signature support

- API to create PSV - Persistent Signature Value
which is used to store signature and other metadata.
- API to reset PSV after signature error.
- API to configure UMR

lib/mlx5: Add signature support

- API to create PSV - Persistent Signature Value
which is used to store signature and other metadata.
- API to reset PSV after signature error.
- API to configure UMR with CRC32C capabilities
- mkey pool can be created with signature capabilities
- spdk_mlx5_device_query_caps reports crc32c support

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Signed-off-by: Sergey Gorenko <sergeygo@nvidia.com>
Change-Id: I6a0fa9a6a1de332381b35f0f7a07350b8814149d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23136
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

lib/idxd: add descriptors for DIX generate

Add DIX generate descriptor and completion record.
Add a new supported opcode to spdk_internal/idxd.

Change-Id: I14bddfaaca2707c52a3e8e0fa416fa770ca0336d

lib/idxd: add descriptors for DIX generate

Add DIX generate descriptor and completion record.
Add a new supported opcode to spdk_internal/idxd.

Change-Id: I14bddfaaca2707c52a3e8e0fa416fa770ca0336d
Signed-off-by: Krzysztof Karas <krzysztof.karas@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24311
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>

show more ...

add new trace BDEV_RAID_IO_START and BDEV_RAID_IO_DONE

Through the both new trace, we can analyze the time consumed by
IO at the RAID layer and the time consumed by each base bdev IO.
This is a trac

add new trace BDEV_RAID_IO_START and BDEV_RAID_IO_DONE

Through the both new trace, we can analyze the time consumed by
IO at the RAID layer and the time consumed by each base bdev IO.
This is a trace example of raid1 IO:

BDEV_IO_START id: i3481
BDEV_RAID_IO_START id: R870 (i3481)
BDEV_IO_START id: i3482 (R870)
BDEV_IO_START id: i3483 (R870)
BDEV_IO_DONE id: i3482 (R870) time: 14.769
BDEV_IO_DONE id: i3483 (R870) time: 15.319
BDEV_RAID_IO_DONE id: R870 (i3481) time: 16.521
BDEV_IO_DONE id: i3481 time: 16.804

Change-Id: I228b660594a39721567e3a0d606b20e8e8b69e1b
Signed-off-by: Xupeng Mingtu <xupeng9@staff.sina.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24894
Reviewed-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

rdma_utils: add spdk_rdma_cm_id_get_numa_id()

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: Id1a39601b23b51605c45a1c6ea817af0152f6819
Reviewed-on: https://review.spdk.io/gerrit/c/spd

rdma_utils: add spdk_rdma_cm_id_get_numa_id()

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: Id1a39601b23b51605c45a1c6ea817af0152f6819
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24511
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <ben@nvidia.com>
Reviewed-by: GangCao <gang.cao@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

sock: remove "socket" from "numa_socket_id"

We will just use "numa_id" to refer to a NUMA socket ID (also
called NUMA node ID). This will eliminate any confusion over conflicts
with "socket" used to

sock: remove "socket" from "numa_socket_id"

We will just use "numa_id" to refer to a NUMA socket ID (also
called NUMA node ID). This will eliminate any confusion over conflicts
with "socket" used to describe TCP sockets.

These APIs were all added since last release, so we can change
this freely.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: Ifb8591c1066a61a78c31ad0e40745a59abb2759f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24606
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <ben@nvidia.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>

show more ...

add new trace BLOB_REQ_SET_START and BLOB_REQ_SET_COMPLETE.

Through the both new trace, the bdev_io trace of lvol and the bdev_io
trace of base bdev of blobstore can be connected in series. Like thi

add new trace BLOB_REQ_SET_START and BLOB_REQ_SET_COMPLETE.

Through the both new trace, the bdev_io trace of lvol and the bdev_io
trace of base bdev of blobstore can be connected in series. Like this:

RDMA_REQ_RDY_TO_EXECUTE id: r31
BDEV_IO_START size: 8 id: i2964 (r31)
BLOB_REQ_SET_START id: a11 (i2964)
BDEV_IO_START size: 8 id: i2965 (a11)
BDEV_NVME_IO_START id: N2957 (i2965)
NVME_PCIE_SUBMIT id: p2959 (N2957)
NVME_PCIE_COMPLETE id: p2959 (N2957)
BDEV_NVME_IO_DONE id: N2957 (i2965)
BDEV_IO_DONE id: i2965 (a11)
BLOB_REQ_SET_COMPLETE id: a11 (i2964)
BDEV_IO_DONE id: i2964 (r31)
RDMA_REQ_EXECUTED id: r31

Change-Id: I607428b094031814627ac4192b9e3defe7ed5337
Signed-off-by: Xupeng Mingtu <jingmamour@gmail.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24130
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>

show more ...

lib/event: Modify spdk_reactor_set_interrupt_mode() to be called from scheduling reactor

The previous behavior of spdk_reactor_set_interrupt_mode()
returning an error when called from a non-app_thre

lib/event: Modify spdk_reactor_set_interrupt_mode() to be called from scheduling reactor

The previous behavior of spdk_reactor_set_interrupt_mode()
returning an error when called from a non-app_thread context
has been changed to return an error when called from a
non-scheduling reactor context. As a result,
spdk_reactor_set_interrupt_mode() must now be executed from a
scheduling core rather than an app thread. Modifications have
been made to address instances where this requirement was not met.

There is no issue when the app thread operates on the scheduling
core, but if it runs on a different core, changing the interrupt
mode will result in an error stating that it can only be modified
on the scheduling core. This could cause scheduling operations
to malfunction. To address this, modifications have been made.

Change-Id: I1f73c84191f5cab877a29405756f106cfb410932
Signed-off-by: syeon.shin <syeon.shin@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24324
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>

show more ...

sock: change socket_id to int32_t

These changes were added since last release, so it's safe to just
change them now without breaking API/ABI.

I mistakenly used uint32_t for the socket_id, when in f

sock: change socket_id to int32_t

These changes were added since last release, so it's safe to just
change them now without breaking API/ABI.

I mistakenly used uint32_t for the socket_id, when in fact DPDK and
our env layer really treat it as an int. This mostly shows up
with the SPDK_ENV_SOCKET_ID_ANY that is defined as -1, which forces
a lot of annoying casting if we're trying to pass this value through
a uint32_t interface.

This socket_id concept will be getting plumbed into
nvme, bdev and nvmf layers, so instead of proliferating the
uint32_t socket_id there, just change sock layer to use
int32_t instead.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: I573ba7f9aeed3b216bef72f39fc671c9c6688d3a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24577
Reviewed-by: Ben Walker <ben@nvidia.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Krzysztof Karas <krzysztof.karas@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>

show more ...

lib/mlx5: Remove spdk_mlx5_crypto_set_attr

THe only user was accel_mlx5, this function
is not used anymore

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I10b947c175f61cd583c6cd814

lib/mlx5: Remove spdk_mlx5_crypto_set_attr

THe only user was accel_mlx5, this function
is not used anymore

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I10b947c175f61cd583c6cd814afb48f7b1cf4b8b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23119
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Community-CI: Mellanox Build Bot

show more ...

accel/mlx5: Use internal PRM based mlx5 library

That gives much higher performance
Other changes
- Add API to get verbs qp from mlx5 qp - needed
for memory domain translation
- Account number of exp

accel/mlx5: Use internal PRM based mlx5 library

That gives much higher performance
Other changes
- Add API to get verbs qp from mlx5 qp - needed
for memory domain translation
- Account number of expected CQ entries to skip busy
polling of CQ
- Count number of operations submitted to QP to prevent
QP overflow
- Use mlx5 mkey_pool API

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I42a2f3305e1d4fa876a674b5c9f572a5802e0847
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23118
Reviewed-by: Ben Walker <ben@nvidia.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

nvmf/tcp: replace pending_buf_queue with nvmf_tcp_request_get_buffers

Use iobuf internal queuing capability exposed via new
nvmf_tcp_request_get_buffers callback mechanism for the TCP requests
that

nvmf/tcp: replace pending_buf_queue with nvmf_tcp_request_get_buffers

Use iobuf internal queuing capability exposed via new
nvmf_tcp_request_get_buffers callback mechanism for the TCP requests
that awaits free buffers instead of queuing them internally.

Introduce new spdk_nvmf_tcp_req_state that clearly indicates if all
requested buffers are allocated for the request.

TCP requests could also wait for the spdk_nvmf_tcp_control_msg_list
buffer for the PDU. Add tcp-only waiting list for such case.

It is the first step in effort to remove pending_buf_queue completely
from the spdk_nvmf_request and rely only on the iobuf and
transport-specific queues.

Change-Id: Ic717d5b8b1090d7703d39f298777d8fdfc30103c
Signed-off-by: Krzysztof Goreczny <krzysztof.goreczny@dell.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23215
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>

show more ...

nvmf/tcp: fix spdk_nvmf_tcp_control_msg_list queuing

control msg buffers are in use when an admin or fabric command has in
capsule data size bigger than the transport->opts.in_capsule_data_size.
Num

nvmf/tcp: fix spdk_nvmf_tcp_control_msg_list queuing

control msg buffers are in use when an admin or fabric command has in
capsule data size bigger than the transport->opts.in_capsule_data_size.
Number of buffers is limited and when there are no free ones request is
queued in the pending_buf_queue.
However, PDU processing in nvmf_tcp_sock_process continues ignoring lack
of the buffer, calling nvmf_tcp_capsule_cmd_hdr_handle() again so
another request is queued for the same PDU, again not having the control
msg buffer available and so on.
Fix this by adding one more PDU processing state that indicates lack of
buffer for the PDU. This way PDU processing continues only after
nvmf_tcp_process() acquires the buffer.

Change-Id: I5e16d7fa3931efc45646f8fdd23794a640c75e93
Signed-off-by: Krzysztof Goreczny <krzysztof.goreczny@dell.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24306
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>

show more ...

lib/mlx5: Update keytag creation to use PRM

The new way to create DEKs is not compatible with
mlx5dv API used by accel_mlx5, so both methods of creating
DEKs remain in this patch. Legacy way will be

lib/mlx5: Update keytag creation to use PRM

The new way to create DEKs is not compatible with
mlx5dv API used by accel_mlx5, so both methods of creating
DEKs remain in this patch. Legacy way will be removed soon.

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I7be82f6ec6cbacc56ab1482537a806792a7e2a77
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23107
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <ben@nvidia.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>

show more ...

lib/mlx5: API to query device capabilities

Only crypto caps are fetched for now, the API will
be extended later

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I0fafe6e4de3a116ed682

lib/mlx5: API to query device capabilities

Only crypto caps are fetched for now, the API will
be extended later

Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I0fafe6e4de3a116ed68226feb28548af8645a722
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23106
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Community-CI: Mellanox Build Bot

show more ...