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 ...

fio/bdev: fix typo in README

Change-Id: Ib0d9e22a1aefca1470c45eb3de53497ece74d2e3
Signed-off-by: Ankit Kumar <ankit.kumar@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/25221

fio/bdev: fix typo in README

Change-Id: Ib0d9e22a1aefca1470c45eb3de53497ece74d2e3
Signed-off-by: Ankit Kumar <ankit.kumar@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/25221
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeliu@tencent.com>
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>

show more ...

fio/nvme, fio/bdev: use specific NUMA ID when possible

Currently if any namespace or bdev reports NUMA_ID_ANY, we allocate
the buffer using NUMA_ID_ANY. This basically keeps original behavior,
unles

fio/nvme, fio/bdev: use specific NUMA ID when possible

Currently if any namespace or bdev reports NUMA_ID_ANY, we allocate
the buffer using NUMA_ID_ANY. This basically keeps original behavior,
unless all namespaces and bdevs reports a NUMA ID.

But it is OK to just use a specific NUMA ID as long as none of the
other namespaces or bdevs in that job have a conflicting NUMA ID. If
the other namespaces or bdevs report NUMA_ID_ANY we can just ignore
them.

Suggested-by: Konrad Sztyber <konrad.sztyber@intel.com>
Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: I18743a5e26d1098804b7783c592b89bfe9286423
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24815
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>

show more ...

fio/bdev: use numa_id when allocating io buffers

fio does the io buffer allocations on a per-thread basis, so if
a thread/job is associated with multiple bdevs, and those
bdevs don't share the same

fio/bdev: use numa_id when allocating io buffers

fio does the io buffer allocations on a per-thread basis, so if
a thread/job is associated with multiple bdevs, and those
bdevs don't share the same numa_id, we will just revert to
SPDK_ENV_NUMA_ID_ANY.

Also, if any of the bdevs report SPDK_ENV_NUMA_ID_ANY, revert
to SPDK_ENV_NUMA_ID_ANY for that case too.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: I92caf0dadc15af3196ae9e0f5ae410ad4cee28f3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24155
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Ben Walker <ben@nvidia.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

fio/nvme: use numa_id when allocating io buffers

fio does the io buffer allocations on a per-thread basis, so if
a thread/job is associated with multiple controllers, and those
controllers don't sha

fio/nvme: use numa_id when allocating io buffers

fio does the io buffer allocations on a per-thread basis, so if
a thread/job is associated with multiple controllers, and those
controllers don't share the same numa_id, we will just revert to
SPDK_ENV_NUMA_ID_ANY.

Also, if any of the ctrlrs report SPDK_ENV_NUMA_ID_ANY, revert
to SPDK_ENV_NUMA_ID_ANY for that case too.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: I2ce39eda613b59e716c1e95c214cce844c507dbc
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24151
Community-CI: Mellanox Build Bot
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <ben@nvidia.com>

show more ...

app/fio/bdev: remove use of spdk_subsystem_init_from_json_config

This function is deprecated, use spdk_subsystem_load_config
instead.

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

app/fio/bdev: remove use of spdk_subsystem_init_from_json_config

This function is deprecated, use spdk_subsystem_load_config
instead.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: I0eb189550706b6f76df5263b3ad9a62f5fe11be9
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24663
Community-CI: Mellanox Build Bot
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

show more ...

env: add opts_size to spdk_env_opts

This will help with ABI compatibility as new fields get added
to this structure.

User is responsible for setting opts_size. Major version has been
bumped, so use

env: add opts_size to spdk_env_opts

This will help with ABI compatibility as new fields get added
to this structure.

User is responsible for setting opts_size. Major version has been
bumped, so users will need to rebuild against latest code, if they
do not set opts_size we will just treat that as if the structure
ends right before opts_size. We will also print an ERRLOG to warn
user to set opts_size appropriately.

As part of this patch, update all in-tree callers of
spdk_env_opts_init() to set the opts_size.

Signed-off-by: Jim Harris <jim.harris@samsung.com>
Change-Id: If2611eccc8b3c43e1dc1654028d000a8a12a544e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23940
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz@tzawadzki.com>

show more ...

nvme/fio_plugin: update the way ruhs descriptors are fetched

For FIO_IOOPS_VERSION >= 35 the fdp backend of fio has removed the
limitation of 128 ruhs descriptors. This way users can specifiy upto 1

nvme/fio_plugin: update the way ruhs descriptors are fetched

For FIO_IOOPS_VERSION >= 35 the fdp backend of fio has removed the
limitation of 128 ruhs descriptors. This way users can specifiy upto 128
placement ID indicies in the range of 0 <= index < max ruhs descriptor
reported by the device.

The is done by allowing FIO to fetch ruhs in two steps. As fio is
unaware of the number of reorted ruhs descriptors, the first call from
fio to the ioengine hook fdp_fetch_ruhs is to figure that out. The
second call from fio is to get all the actual ruhs descriptors.

Change-Id: I1026315e14c539f397e9c3d4a3bc6addf53b04c1
Signed-off-by: Ankit Kumar <ankit.kumar@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24238
Community-CI: Mellanox Build Bot
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>

show more ...

nvme/fio_plugin: add support for the disable_pcie_sgl_merge option

Add support for the disable_pcie_sgl_merge option to the fio nvme
plugin. This allows users to disable merging of physically contig

nvme/fio_plugin: add support for the disable_pcie_sgl_merge option

Add support for the disable_pcie_sgl_merge option to the fio nvme
plugin. This allows users to disable merging of physically contiguous
SGL elements for the PCIe transport.

Change-Id: Ib4fda74f0ffc18b83d71cdb154534fccfeda487b
Signed-off-by: Vincent Fu <vincent.fu@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/24195
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>

show more ...

fio_plugin: move current_f to spdk_fio_probe_ctx

Now that we have a probe context structure, move current_f there, as
it's more fitting place to keep it.

Signed-off-by: Konrad Sztyber <konrad.sztyb

fio_plugin: move current_f to spdk_fio_probe_ctx

Now that we have a probe context structure, move current_f there, as
it's more fitting place to keep it.

Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I101e0ea074077c7779ad5d7ac225f3b20820a357
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23589
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>

show more ...

fio_plugin: set hostnqn via trid

Host NQN can now be set in two ways - using the global option or via
trid in the filename parameter. The latter takes precedence and will
overwrite the global optio

fio_plugin: set hostnqn via trid

Host NQN can now be set in two ways - using the global option or via
trid in the filename parameter. The latter takes precedence and will
overwrite the global option, if set.

That way, it's possible to specify different host NQNs for each job and
it makes host NQN configuration more similar to other tools (e.g.
spdk_nvme_perf).

Fixes #3397

Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I60e7264e5de42eafe2d195c3069ab4dddfeb1a06
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/23588
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <jim.harris@samsung.com>

show more ...

nvme/fio_plugin: trim add support for multiple ranges

For FIO_IOOPS_VERSION >= 34 fio supports multiple ranges per trim
request. fio option num_range can be used to specify that.
If multiple ranges

nvme/fio_plugin: trim add support for multiple ranges

For FIO_IOOPS_VERSION >= 34 fio supports multiple ranges per trim
request. fio option num_range can be used to specify that.
If multiple ranges are specified the io_u data buffer contains the
start offset and number of bytes to be trimmed for each range.

This commit adds the necessary changes to support multi range trim by:
* allocating per IO request dsm range buffer with correct size.
* parsing the io_u data buffer and creating the trim request.

Change-Id: Ied274feeaba6dc36cab157ffd217730f9560da73
Signed-off-by: Ankit Kumar <ankit.kumar@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/22890
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>

show more ...

nvme/fio_plugin: add trim support

Add dsm_range per IO request for trim (DSM deallocate).
Spdk supports asynchronous trim requests for all cases except ZBD where
trim does a zone reset. Set ioengine

nvme/fio_plugin: add trim support

Add dsm_range per IO request for trim (DSM deallocate).
Spdk supports asynchronous trim requests for all cases except ZBD where
trim does a zone reset. Set ioengine flag FIO_ASYNCIO_SYNC_TRIM in that
case.
While queuing add the necessary changes to fill dsm_range and submit the
request.

Change-Id: Ie7971c7cdf622d3fc44fe057c1b8991a1909e0e9
Signed-off-by: Ankit Kumar <ankit.kumar@samsung.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/22889
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>

show more ...

test: ensure unique address for FIO plugin

FIO plugin (fio/spdk_bdev) always starts RPC server on
SPDK_DEFAULT_RPC_ADDR address. This is unnecessary, because in most
cases there will be no need for

test: ensure unique address for FIO plugin

FIO plugin (fio/spdk_bdev) always starts RPC server on
SPDK_DEFAULT_RPC_ADDR address. This is unnecessary, because in most
cases there will be no need for RPC interface to be active for
communication. We can change plugin's implementation similarly to what
we have in bdevperf: RPC server is going to be started only when
'--spdk_rpc_listen_addr' is specified.


Change-Id: I99227626c71ae5615cbdd5090fcdd4c42f193eaa
Signed-off-by: Krzysztof Karas <krzysztof.karas@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/22682
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Reviewed-by: Boris Glimcher <Boris.Glimcher@emc.com>

show more ...

fio: Promote bdev fio plugin to app directory

We are going to start installing this in the near future.

Change-Id: I98ddf5fadf7505dfae2689ef4df54e7e0bf10f8c
Signed-off-by: Ben Walker <ben@nvidia.co

fio: Promote bdev fio plugin to app directory

We are going to start installing this in the near future.

Change-Id: I98ddf5fadf7505dfae2689ef4df54e7e0bf10f8c
Signed-off-by: Ben Walker <ben@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/21389
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot

show more ...

fio: Promote the spdk_nvme fio plugin to app

In the future we will install this.

Change-Id: I5fbeb87b897b90df662c30784b8857b334ddfb2f
Signed-off-by: Ben Walker <ben@nvidia.com>
Reviewed-on: https:/

fio: Promote the spdk_nvme fio plugin to app

In the future we will install this.

Change-Id: I5fbeb87b897b90df662c30784b8857b334ddfb2f
Signed-off-by: Ben Walker <ben@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/21388
Tested-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <jim.harris@samsung.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>

show more ...