There's one left has not been changed.
That's be ignored ever.
Change IO/s to IOPS.
Signed-off-by: dongx.yi <dongx.yi@intel.com>
Change-Id: I1555cc02644b238e8d4fece31f932a480c20121d
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/478399
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
'delay_pcie_doorbel' parameter in 'spdk_nvme_io_qpair_opts' structure
was renamed to 'delay_cmd_submit' to make it suitable for every
transport. Old name is also kept for backward compatibility.
Signed-off-by: Evgeniy Kochetov <evgeniik@mellanox.com>
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: I09ef8028133c4a3d4a5bbc5329ced1f065bcaa46
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/475305
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>
Community-CI: Broadcom SPDK FC-NVMe CI <spdk-ci.pdl@broadcom.com>
There was an assignment in perf that prevented the user from actually
disabling keep_alive. If the user specified 0, it would get overridden
by the default opts value. So set the global in perf to a reasonable
value, and always use that value when setting the keep alive timeout.
Change-Id: I2eba14525613d679d7def29cc9ed312851cae156
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/473059
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>
This is sort of like the periodic stats that fio
prints every second or so. We can add more to this
later, but this by itself is a nice improvement I
think.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I4ac588c8301bae9924dd6541cbc386ffe8d06b5d
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/473339
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>
This will be used in a future patch.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie88daa20842f5f166b4a7a5abccafd6683d6da64
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/473338
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Alexey Marchuk <alexeymar@mellanox.com>
There's little use to the "Starting thread on core XX"
messages - they may have served a purpose at some point,
but the other messages showing where each controller
is associated is sufficient.
Also remove the "Attaching to..." message - the
"Attached to..." message is sufficient here too.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9543e8f3f6d7b6e28238cba27fc1e4194908f3b7
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/473337
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Alexey Marchuk <alexeymar@mellanox.com>
The core number may be 1 or 2 characters, and the
printf format wasn't accounting for that.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I3c87466df1014fb6c0a4e7db1a9b6f6054559eae
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/473065
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Alexey Marchuk <alexeymar@mellanox.com>
When we specify a queue depth that may be higher than
the io_queue_size, we print a warning indicating this.
But it was also setting the g_warn flag which would
result in this message also getting printed:
WARNING: Some requested NVMe devices were skipped
But we don't actually skip any SSDs in this case. So
remove setting of that flag in this case.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ic63209f1a96990dac9664ce245c6406c37248494
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/473064
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
The puprose is to print some trace info. We support
it in nvme identitfy. Also -L is occupied by other usage,
so we use -T instead.
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: I29f018c2a5951e5e3aad04dbeaa7ee74a2d7e3f4
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/469919
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
When PRACT is set, if metadata size is 8 bytes, PI is stripped
(read) or inserted (write). Hence block size must not include
metadata size for extended LBA payload. This patch fixes the issue
by reducing metadata size from block size for this case.
On the other hand, When PRACT is set, if metadata size is larger
than 8 bytes, PI is passed (read) or replaced (write). So block
size is not necessary to change for this case.
The wrong block size didn't cause any visible issue but should be fixed.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I3351e3e3b7816f726752e85604cf557251d9870c
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/468018
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
g_metacfg_prchk_flags had been used to pring PRACT setting. This
didn't cause any issue but will lead to misread.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I2f6c48b12ffa511afa2663d31dc9467cec980c9c
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/468017
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
perf is under the examples directory, but it's much more than
an example application and probably needs to be moved. We use
the libaio functionality extensively as part of benchmarking
and comparing the SPDK driver with the kernel driver. So we
need to keep this functionality.
This reverts commit b3d9ebae21.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ifb9a1df919d32a98c328101029cc22e91915a976
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/466795
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Users can use '-N' option to disable shutdown notification
for faster shutdown process.
Change-Id: I227dc4da3aa2f89d23f9b235e979cb6d3e25d635
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/463950
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>
Original intention most likely was to compare nvme lib to
aio, but since then aio bdev was developed.
There should be no need to keep the support for AIO in
perf example app.
Change-Id: I716f2164b8eaef16db05b4fffa35393c7d916e0b
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/461593
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
New flag -V was added for VMD enabling.
Change-Id: Ie0feae1bdfce39ad246f57411d00587c6417d21d
Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/450926
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Data offset are intended to correspond to DATAO in NVMe/TCP and
Buffer Offset in iSCSI.
Previously for iSCSI, buffer offset had been merged to start block
address, but passing buffer offset separately from start block address
clarifies the logic more.
On the other hand, for NVMe/TCP, passing DATAO separately from start
block address will be critically important because DATAO will bave any
alignment and will be necessary to use for not only reference tag
but also guard computation.
This patch adds data_offset to struct spdk_dif_ctx and adds it to the
parameters of spdk_dif_ctx_init(). ref_tag_offset is also added to struct
spdk_dif_ctx and it is computed by dividing data_offset by data_block_size
and is used to compute reference tag.
The next patch will use this change when getting DIF context in SCSI.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Id0e12ca9b1dc75d0589787520feb0c2ee0f844a5
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/457540
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>
verify_io is a function pointer, so avoiding the
call altogether is beneficial. We know that no
verification is possible if md_size == 0, so use
this to determine if the call should be made.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I75f01e4f56aacb0f7babd2342efc6014f886667d
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456249
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
The compiler wasn't always inlining this. Inlining
results in a small optimization.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I34240d85439b7329fab7151ae4f8b02f44c4d7d5
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456248
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Now that we print the addr/nqn for fabrics controllers,
adjust the column width in the output so that it prints
the full information for fabrics, but keeping a smaller
width when it is only PCIE controllers.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I3532a8928a806b8084047f6686c37cd02d015983
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456079
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
The serial number doesn't provide much value - but the
transport id can sometimes be helpful though to correlate
performance of different devices to their PCI BDF.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I529f111311108a6efc1ba23129963bb4383d3bdd
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456078
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
These are the names displayed in the output. Some
upcoming patches will modify these names, so making
this a common function now will reduce the work
later.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I11cb8f607a66523becdcf621d8d6c911aec4427d
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/456077
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
The perf tool's bandwidth data is MiB/s, not MB/s.
So update the data description part.
Change-Id: I770cc0d7c0f0a4d56cb4eff593e88fbed55e3ed6
Signed-off-by: Liang Yan <liang.z.yan@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/451319
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
For some testing, we need queue pairs to exist but not actually
be in use.
Change-Id: I2b17ff0172c9ec002692babcf7d4d612c3062eb4
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/392977
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Queues will be doled out to cores as available.
Change-Id: Ib6a0fe846a9d90b659754be1c11ae022abbe38a3
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/391876
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>
This tool continually polls for completions, so it
is safe to turn this on.
Change-Id: Ice1c68cdaff070f8edd428621e19a6fb44fb8c31
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447712
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>
Allow user to add seed value for guard compuation to DIF context.
This will avoid the guard being zero in case of all zero data.
NVMe controller doesn't support seed value for guard computation
explicitly, and hence if we want to use such a seed value in
NVMe controller, we have to format metadata more than 8 byte,
and add seed value into the reserved metadata field.
But some popular iSCSI/FC HBAs and SAS controllers have supported
seed value for guard computation, and so supporting seed value
in the SPDK DIF library is very helpful for some use cases.
Hence this patch makes the DIF library possible to specify seed
value for those use cases.
Change-Id: I7e9e87cb441bf263e64605c7820409fdc22dd977
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/444334
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: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: wuzhouhui <wuzhouhui@kingsoft.com>
If do not do that, we will continue polling. And checked
the exceptional handling of aio_check_io, it will just exit.
So we add such case, thus we shall not poll the qpair again.
This is especially useful when we use perf to test against
NVMe-oF target.
Change-Id: Ib820e0b80f80cfceb1ea5e08b359f50d9e360b30
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/c/444325
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
If spdk_nvme_probe fails, there will be no controller,
and we jump to clean up. And the thread is not created,
so we need to judge the value of thread. if it is not zero,
we will do the pthread_cancel. Otherwise, we do not
need to call that. If we call that, it will have coredump.
btw: we also fix an small issue of cleanup.
Change-Id: Iee854f6ef42ef6d4a3d3f63d57740375e293a3ba
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/c/444147
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
perf application can't generate IO for NVMe namespace with
more than 4G size.
Example of error:
"Attached to NVMe over Fabrics controller at 1.1.75.1:1023:
nqn.2016-06.io.spdk.r-dcs75:rd0
WARNING: controller SPDK bdev Controller (SPDK000DEADBEAF00 ) ns 1 has
invalid ns size 0 / block size 4096 for I/O size 4096
WARNING: Some requested NVMe devices were skipped
No valid NVMe controllers or AIO devices found"
ns_size variable is uint32_t, spdk_nvme_ns_get_size function
returns uint64_t. Result can exceed the maximum size of
uint32_t and ns_size remains 0.
The issue introduced by commit: f2462909
Change-Id: Idc6dd8688d5d6268bda1a1d6b06a611643af6155
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Reviewed-on: https://review.gerrithub.io/c/443996
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>
Keep alive was enabled with RDMA/TCP transport by default, for
perf tool we should also enable it or the nvmf target may
disconnect when timer expires.
Fix issue #609.
Change-Id: Ibd4073c4a4ac398d601285d3bf64631058046c59
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/441834
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@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>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This patch adds DIF generation and verification for NVMe namespace
formatted with DIX.
Change-Id: Ic67c88c485aacffa40fd0c73ba164f2cb1f88991
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/439969
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>
For the commands sent from upper layer, the NVMe driver may
split one command into several children requests based on
stripe and data length, so number of requests in each queue
pair has a value which is start from 512. we must ensure
that the number of request in each queue pair is big enough
to process all the commands, e.g: a user's input for a 512KiB
read can be divided into 5 children NVMe commands when maximum
transfer length is 128KiB, in addition, one parent request is
occupied to track children commands.
Fix issues #566 and #573.
Change-Id: I162da8e1e15692625ce311e68a72c89b6492dd56
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/440457
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>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Apply the new DIF library to IO submission and verification.
Condition to use DIF generation and verification are not changed
in this patch.
appmask, apptag, and lba in struct perf_task are duplicated with
members of struct spdk_dif_ctx and are removed in this patch.
Change-Id: I68ccfb41cce2546d7145417c1490f49f0e62bb7a
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/437907
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>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Currently, DIF information is extracted by NVMe name space API each
time. This patch changes to extract and hold DIF information at
namespace registration and use it.
Change-Id: I6ce857bf5df5fd5972bdca8b52226d1fb81e40c7
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/439653
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>
Remove unused apptag_mask and apptag in struct ns_entry.
Change-Id: I2da7ee099c8b18c93a9525e7195c1f98bc5a9543
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/440674
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
This patch replaces buf by iovec of struct perf_task.
SGL is not used in AIO and NVMe IO yet but SGL is necessary to
handle DIF and DIX by using the new DIF library.
Using SGL in AIO and NVMe IO will be done later.
Change-Id: Ied8f2aa0cb9cd933986e8046df7d48666bcc2f89
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/440673
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Introduce a function pointer table to ns_entry to remove if-else
sequence in every operation depending on the type, AIO or NVMe.
This will simplify upcoming DIF and DIX support in the Perf tool.
Change-Id: Ibbd9a9ac3a0b5df529d5b60706ce750a746114c3
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/439630
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Subsequent patches will introduce function pointer table for
IO type depenedent operations.
This patch doesn't cause any functional change and just tries to
make them easier.
Change-Id: Ib612be43b9cf4bbb620739b29913cd05151bc872
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/439805
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: wuzhouhui <wuzhouhui@kingsoft.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Subsequent patches will introduce function pointer table for
IO type depenedent operations.
This patch doesn't cause any functional change and just tries to
make them easier.
Change-Id: I664c9e39d101957c55cfcf9426f82feca348c328
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/440670
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: wuzhouhui <wuzhouhui@kingsoft.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Perf tool had checked status of each NVMe IO command submission but
had not checked completion status of each NVMe IO command processing.
This patch checks the completion status of each NVMe IO command
processing and prints error if found.
No error handling is not implemented yet but this addtion will be
of any help for subsequent DIF patches.
Change-Id: I8da52d97584d7688cff04092efee658556cf7d86
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/439966
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: wuzhouhui <wuzhouhui@kingsoft.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Subsequent patches will support DIF and DIX in NVMe Perf and
will make buffer management a little complex. This patch tries
to make them a little easier.
Change-Id: I3e0e3e03ca386467c7477e1ec8aa537ca47316e2
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/437903
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
If user want to set PRCHK, current Perf tool requires user to
add ',' as a prefix to PRCHK.
This patch removes the limitation by referring /lib/nvme/nvme.c
Change-Id: I327810b38c02116b9580873b6046bdca55b5162a
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/c/437913
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
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>
This patch is used to make the io completion
of different ns_ctx owned by each worker in
a round robin way.
Purpose: To avoid the timeout if there are many
ns_ctxes. For example, if each ns_ctxt connects
to remote NVMe-oF target with qpair. If there are
many qpairs, we may face the time out of some qpairs.
Change-Id: I477cb7436dc46ea498a26f990ed79001fa1bf2d6
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/c/438150
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>
There are 2 #include "spdk/env.h" and an extra "]".
Change-Id: I7003333a164ef38ed166fbecf9e6b9f2df350ad9
Signed-off-by: WangHaiLiang <hailiangx.e.wang@intel.com>
Reviewed-on: https://review.gerrithub.io/416373
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@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>
Add init_crc parameter as seed value to spdk_crc16_t10dif API to generate
a CRC value spanning multiple separate buffers.
This will be necessary for upcoming DIF/DIX patches.
Having init_crc parameter is general, and so change the existing API
without adding seed version of the existing API.
Change-Id: I0ac7919b18013967e41829dcedd3e4e73204d5d6
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/437204
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: Ia65e235a85207c128ba274e1bab38d6c35344239
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/435563
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Format of namespace is not changed during running but namespace
information is collected by identify command for every IO.
This caused much degradation of performance.
Hence call spdk_nvme_ns_get_data() and hold the pointer to the
buffer in which namespace information is stored at registration
of namespace and use the pointer in both task_extended_lba_setup_pi()
and task_extended_lba_pi_verify().
Change-Id: I12461ccb763247bbb9ccbe34802805ace67d9400
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-on: https://review.gerrithub.io/424362
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@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>
'-s' option is usually associated with the memory size,
but our ioat/perf and nvme/perf use it for io size.
To keep those consistent:
'-s' option (io size) was be changed to '-o'
'-d' option (mem size) was be changed to '-s'
Change-Id: I97cef4c0a4e6e902277dd50dfb43bde420a6427e
Signed-off-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/422654
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Transport addresses can now additional contain an "ns"
key whose value is a specific nsid to connect to. If
unspecified, all namespaces will be used, which is
the previous behavior.
Change-Id: Icbebe32d68a1f97acaa1197c7fec5f119f33ec4d
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/421267
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@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: Xiaodong Liu <xiaodong.liu@intel.com>
Low level NVMe driver will set the hardware IO queue size
based on MQES and users' input, perf tool has requested
maximum IO queue size when started, but for the number
of IO requests, it uses the default value. While here,
set the number of IO requests to a higher value, so that
for high queue depth tests, the perf tool will not fail.
Fix the issue #256.
Change-Id: I6d11ec65e0f3b246b824875fa17d74a31fd2702b
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/403256
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
- Add support for multi page CNS 0x2
- Use CNS value 0x02 (SPDK_NVME_IDENTIFY_ACTIVE_NS_LIST)
to query active namespaces
- Add an API to iterate the active list
Change-Id: Iea524881fa6e3610a7d85ab02a2005a92fd633df
Signed-off-by: John Meneghini <johnm@netapp.com>
Reviewed-on: https://review.gerrithub.io/401957
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The vendor:device ID is already printed out in attach_cb, so printing it
again in probe_cb isn't that useful.
This also lets us remove the last use of spdk_pci_get_device(), allowing
it to be deprecated and removed in future patches.
Change-Id: Ie65abf136cce4d413f06594f91533ab9f9d4aca7
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/405708
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This allows us to remove most uses of spdk_pci_get_device(), which looks
up a PCI device structure from an arbitrary PCI address. This function
is problematic, since it uses internal DPDK data structures that aren't
meant to be part of the public API. There is still one use in the
codebase, which will be cleaned up in another patch.
Change-Id: Ia1fe1f799c240195f6871c1d92821074f884c4e6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/405707
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Automatically detect more whitespace errors.
All existing cases are fixed; only whitespace change (verify with
diff -w) except for one comment style fixup in include/spdk/nvme.h.
Change-Id: If750e54b9c8e3421ea6feda5f20184a31431631e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/402360
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
For end-to-end data protection tests with extended lba format, when
Protection Information Action bit was cleared to 0, the perf
application can do some comparison based on the flags we used
for writes. For the unexptected protection information value, print
a notice for users.
Change-Id: I9851ce6338030f9b57180cd5e46368e0c94bc212
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/394158
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
This patch is used to fix the time caculation for
average, min, max latency.
Also changes the float to double to make it more accurate.
Change-Id: Ia91278f04656787e2dcf94dd37d839bcf31d618d
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/394674
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Modifies spdk_env_init() and spdk_mem_map_init() such that
they return on failure instead of terminating with exit()
or abort().
Change-Id: I054c1d9b2e46516ff53d845328ab9547f54bdbc4
Signed-off-by: Lance Hartmann <lance.hartmann@oracle.com>
Reviewed-on: https://review.gerrithub.io/393987
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Currently, each histogram range is hardcoded to
128 buckets (1ULL << 7), resulting in 58 ranges
(64 - 7 + 1) and a total size of 58 * 128 *
sizeof(uint64_t) = 59392 bytes.
To allow for more usage models in cases where this
size is prohibitive, enable the following changes:
1) specify number of buckets per range (in number of bits)
2) specify max datapoint value (in number of bits)
The latter can be useful for cases where datapoints
are never close to requiring all 64 bits - and allows
reducing the number of ranges. Any data points that
exceed the max will be tallied in the last bucket
in the last range.
Testing shows no performance disadvantage using the
dynamic approach.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I5979bcdff6209faaa9dee293918ef2a78679bcd4
Reviewed-on: https://review.gerrithub.io/392707
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: I7174f1799361b8337ff5590b90ad6a0564ca8e9b
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/391899
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I127279ba32388bcedb85ed0a2eba9b5345c02eda
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/366532
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
There is a case in high queue depth situation and the IO may be
failed due to no memory. Our perf tool did not recognize this
and always count the IO in the inflight IOs. Thus in the completion,
it waits more than actual IOs to return and the Perf is hanging.
Change-Id: Ie5472e260c75c5abd80c06eb6c613dc7d1761c78
Signed-off-by: GangCao <gang.cao@intel.com>
Reviewed-on: https://review.gerrithub.io/391582
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Require braces around all conditional statements, e.g.:
if (cond)
statement();
becomes:
if (cond) {
statement();
}
This is the style used through most of the SPDK code, but several
exceptions crept in over time. Add the astyle option to make sure we
are consistent.
Change-Id: I5a71980147fe8dfb471ff42e8bc06db2124a1a7f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/390914
Reviewed-by: <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie7282eb2e9f403ae51c449389216bcaf45808277
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/386190
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: I9ced263535e8cd9c9503a208a77832140833c9d3
Signed-off-by: Liang Yan <liang.z.yan@intel.com>
Reviewed-on: https://review.gerrithub.io/383952
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
For perf tests with large I/O size, the NVMe driver may split one
request to several entries based on stripe and maximum transfer size,
so take user's input of queue depth as default queue size of NVMe
queue pair is not accurate. Here the patch evaluate users' input
and report an error when queue entries more than hardware's limit.
Change-Id: I66922a3d673dc97796d7fbe6e86cf5037a45b37d
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/379969
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Change-Id: Ie70236f6be40ad9598969da2a462af44763c4ac4
Signed-off-by: Liang Yan <liang.z.yan@intel.com>
Reviewed-on: https://review.gerrithub.io/379437
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Fully port the remaining uses of rte_ functions from DPDK to the SPDK
env library abstraction layer.
This also simplifies buffer allocation: each task only needs to be
allocated once during the initial submit_io() call, rather than using a
mempool to get/put the task on every I/O.
Change-Id: I39c8caff81bbb1467101ba3b24a389c437075c61
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/378220
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Add a new struct spdk_nvme_io_qpair_opts to allow the user to override
controller options on a per-I/O qpair basis.
Existing callers with qprio == 0 can be updated to:
... = spdk_nvme_ctrlr_alloc_io_qpair(ctrlr, NULL, 0);
Callers that need to specify a non-default qprio should be updated to:
struct spdk_nvme_io_qpair_opts opts;
spdk_nvme_ctrlr_get_default_io_qpair_opts(ctrlr, &opts, sizeof(opts));
opts.qprio = SPDK_NVME_QPRIO_...;
... = spdk_nvme_ctrlr_alloc_io_qpair(ctrlr, &opts, sizeof(opts));
Change-Id: I8ac3ea369535cfde759abbe75e1d974b6450a800
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/369676
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I1aee7aba522cc816f69709cfc95d12c50a5d0f4b
Reviewed-on: https://review.gerrithub.io/367279
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: Ibbe601489d16a9585e56de1c95fe31e9a602a7e0
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/366387
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Remove dpdk_ prefix in spdk_app_opts and spdk_env_opts
Change-Id: I6f231f67072b808e84945d41b1fe31a180beb350
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/365787
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: I7ccff922465195c7fe9836633196cd7a8816c11c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/365071
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
In this patch, we also update perf and identify
examples. If there is no local nvme device info
parsing, we will set dpdk initialization with no-pci
choice.
Change-Id: I58b2d291b7b53894aeb194a16798ff1c72cf25b4
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/365361
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia0083365b2da63cb38aebb9f7bbc02f4dfd1ae94
Reviewed-on: https://review.gerrithub.io/365263
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
DPDK's use of getopt() needs special handling of the optind global
variable since we are passing it a separate array of arguments (not the
typical argv and argc). Set optind to 1 internally to env_dpdk so that
the apps don't need to know about it, and restore optind in case the
calling app is also using getopt().
Change-Id: Icbf07002c99fa9f94c866e8eff707124b0ef679b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/365062
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Also provide an option in perf tool let users to
disable it.
Change-Id: If4952513d77cecaa4f9403fbea811d86916ee87c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/363311
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Gracefully handle ns_count == 0 in print_performance() rather than
asserting.
Change-Id: If8f8d56a2dd4d21ddc61069555c2b90d027431f4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/363614
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: Ben Walker <benjamin.walker@intel.com>
- rename spdk_malloc_socket to spdk_dma_malloc_socket
- rename spdk_malloc to spdk_dma_malloc
- rename spdk_zmalloc to spdk_dma_zmalloc
- rename spdk_realloc to spdk_dma_realloc
- rename spdk_free to spdk_dma_free
Change-Id: I52a11b7a4243281f9c56f503e826fd7c4a1fd883
Signed-off-by: John Meneghini <johnm@netapp.com>
Reviewed-on: https://review.gerrithub.io/362604
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
A single -L can be used to get the latency summary.
Two -L's (or -LL) can be used to get both the latency
summary and the detailed histogram.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I3fc0f4e2dfff7b041a665fe35aa33f11e4c3ebad
Reviewed-on: https://review.gerrithub.io/362270
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I0c47f2086d4f895cd75f32efc7df30d7182adcb1
Reviewed-on: https://review.gerrithub.io/362269
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
The latency tracking is done with ranges of bucket arrays.
The bucket for any given I/O is determined solely by TSC
deltas - any translation to microseconds is only done after
the test is finished and statistics are printed.
Each range has a number of buckets determined by a
NUM_BUCKETS_PER_RANGE value which is currently set to 128.
The buckets in ranges 0 and 1 each map to one specific TSC
delta. The buckets in subsequent ranges each map to twice
as many TSC deltas as buckets in the previous range:
Range 0: 1 TSC each - 128 buckets cover deltas 0 to 127
Range 1: 1 TSC each - 128 buckets cover deltas 128 to 255
Range 2: 2 TSC each - 128 buckets cover deltas 256 to 511
Range 3: 4 TSC each - 128 buckets cover deltas 512 to 1023
Range 4: 8 TSC each - 128 buckets cover deltas 1024 to 2047
Range 5: 16 TSC each - 128 buckets cover deltas 2048 to 4095
etc.
While here, change some variable names and usage
messages to differentiate between the existing latency
tracking via vendor-specific NVMe log pages on Intel
NVMe SSDs, and the newly added latency tracking done
in software.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I299f1c1f6dbfa7ea0e73085f7a685e71fc687a2b
Fix up all existing spacing errors in comments and add an automated
check for patterns like /*comment*/.
Change-Id: I28f61c93612dc0f8aed66bd509da78e91ea9737e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is required by libaio. Previously, buffers were aligned to 512
bytes, but 4K devices need 4K-aligned buffers.
Change-Id: I96080e72dc77e0e72f426f7c9fe98b6724f66e1b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Several examples have a function to associate workers
with threads. Simplify that algorithm.
This seems to just shift some of the complexity
from register_workers down to main, but in the long
run the DPDK threading will get abstracted into
env as well and greatly simplify that part.
Change-Id: Ic106dde58fa5351a1ce0a058161b08062e121d3b
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Since we know the queue depth that we will be using during the test,
request that as the queue size when attaching to NVMe controllers.
Reserve one extra queue entry above the expected queue depth since NVMe
queues must always have one entry free to distinguish between queue
empty and queue full cases.
Change-Id: I809982207edb4894148aec09b10c4e2de4a040d3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
By default, all SPDK applications will not share memory.
To share memory, start the applications with the same
shared memory id.
Change-Id: Ib6180369ef0ed12d05983a21d7943e467402b21a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
