Commit Graph

8466 Commits

Author SHA1 Message Date
Konrad Sztyber
441431d228 lib/trace: don't format arg's name while registering a tpoint
The arguments of a tracepoint are formatted when they're printed now, so
there's no need to append ":" or pad it with spaces.

Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I74f5568f1982dacc079e3b80bd19a9cd740b48ce
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7955
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-06-10 10:58:05 +00:00
Konrad Sztyber
462eb754a2 lib/thread: trace (get|put) IO channel calls
The traces record calls to spdk_(get|put)_io_channel() and saves the
reference count of the IO channel and its context.  The context, instead
of an IO channel pointer, was selected because the same pointer is often
used in other traces (e.g. nvmf's poll group), so it makes it possible
to match these traces together.

Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I15fe982a89685d8f6e23d406d6d48f5c2d9d604b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7232
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-06-10 10:58:05 +00:00
Konrad Sztyber
62aa8bd8d2 nvmf/rdma: trace qpairs instead of rdma_cm_id
Both values should provide similar information, while the qpair can also
be matched to the traces from lib/nvmf allowing the user to track the
qpairs across these modules.

Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Iba9abdd3f41b93100c0403b1c90fc4549d39189e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7159
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-06-10 10:58:05 +00:00
Konrad Sztyber
a266b6e4be nvmf: add traces around poll group lifecycle
The traces are tracking the lifecycle of a poll group: creating it,
adding and disconnecting qpairs, and finally destroying the group.

Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I075b7f24d14b8fbb42bb18ddd70a668a8bace118
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7158
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-06-10 10:58:05 +00:00
paul luse
db71eb6a4b lib/accel: add suport for batching copy + CRC32C command
Future patches will add vectored versions and accel_perf support.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I9cea07283bba871c6d026f8c61c2c9552eaaa23d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8147
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>
Community-CI: Mellanox Build Bot
2021-06-09 07:26:00 +00:00
paul luse
cfd37e28f1 lib/idxd: add support for batching copy + crc32c
Upcoming patches will add accel_fw support for batching this cmd
and then vectored versions later along with accel_perf to exercise
them.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I59f577f4365fbf063d7419cc6052e10c998b58bc
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8143
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-06-09 07:26:00 +00:00
paul luse
a6c5480f1d lib/idxd: fix batch submission of crc32c
Was using reserved field. Similar fix to what was done earlier
for direct submission of crc32c operation.

fixes #1972

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: Ie9867e72f60c7f38aa1af0273a036f34580ed4c6
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8145
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Community-CI: Mellanox Build Bot
2021-06-09 07:26:00 +00:00
paul luse
fa99984283 lib/idxd: refactor batch prep command function
To match regular sumission prep function and allow caller to
modify both descriptor and completion structures.  Also allows
for more accurate error reporting.  Needed for upcoming patch to
fix batch CRC submissions.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I4b7b5b24e2f54149b513d4b23ba32f3802aff3e5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8144
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-06-09 07:26:00 +00:00
paul luse
be8f5a3963 lib/accel: Add suport for copy + CRC32C to accel framework
Upcoming patches will add support for the idxd engine and
the accel_perf tool.  Also following will come vectored support and
batch versions.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I7223517a844525ad52ed49d65627b04c3cd9fe7c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8141
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>
2021-06-09 07:26:00 +00:00
paul luse
1076008e63 lib/idxd: add support for new opcode to low level library (copy + CRC)
Upcoming patches will add support to the accel fw, the idxd engine and
the accel_perf tool.  Also following will come vectored support and
batch versions.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: Ie0fbd4b8da9f727426000898c0b511587adac65b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8139
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-06-09 07:26:00 +00:00
MengjinWu
ab5cc9dd6e lib/nbd: refine _nbd_fini process
_nbd_fini will make all NBDs into closing state.
remove _nbd_async, beasue it will call asynchronous error.

Signed-off-by: MengjinWu <mengjin.wu@intel.com>
Change-Id: Ifb873b7f079b735983bdf20c2df652be0a21919f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8035
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-06-09 07:25:44 +00:00
MengjinWu
628c230de4 lib/nbd: check if nbd is registered when start
Signed-off-by: MengjinWu <mengjin.wu@intel.com>
Change-Id: Ie6918e442724020e2be8f98cab971e1126ebb20a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8033
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: GangCao <gang.cao@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Community-CI: Mellanox Build Bot
2021-06-09 07:25:44 +00:00
Changpeng Liu
5c680e29ea nvme/vfio-user: use the common request submission function
Change-Id: Ic2f6dedd0734373d23e80f7875640fb17db7b321
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8032
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2021-06-09 07:21:31 +00:00
Changpeng Liu
d4844d5b4e nvme/pcie: use nvme_pcie_vtophys in the submit request function
And for some internal functions we need to pass controller
parameter so that we can do vtophys based on transport type.

Change-Id: I3ca4fa162ec9305f62b295ba21f7474c21edfe52
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8031
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2021-06-09 07:21:31 +00:00
Jim Harris
6564bd94ba nvme: fix DEFAULT_IO_QUEUE_SIZE_FOR_QUIRK handling
We need to wait to process this quirk until after we
have a valid CAP register value.  Before this fix,
controllers with this quirk would get their io_queue_size
always capped at 2 (min io queue size) because CAP hadn't
actually been read yet.

Fixes: f5ba8a5e (nvme: add NVME_CTRLR_STATE_READ_CAP)

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I4df87b5dfb0faa21db5b4cf6fc667d80621d1691

Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8211
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
2021-06-09 07:07:50 +00:00
Changpeng Liu
5e4fe0adc8 nvme: inline nvme_pcie_vtophys
The inline function can also be used in the coming submit request
function.

Change-Id: If4a5511001e6586dbce0978298beddc537f54d8b
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8173
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-06-08 08:28:44 +00:00
Changpeng Liu
b69827a394 nvme/pcie: move nvme_pcie_qpair_submit_request() to pcie common layer
The PCIE and VFIOUSER both can use this function, the only difference
is VFIOUSER should use IOVA=VA to do the vtophys translation, so
here we will move the function to the common PCIe layer as the first
step.

Change-Id: I699edb67a00a2fa534072fc02ac2dd4a27aba8f4
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8030
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-06-08 08:28:44 +00:00
tyler_sun
4f4f505c77 nvme: get changed ns log once AER notice of ns changed received.
Each time the following file
"/sys/kernel/config/nvmet/subsystems/nqn_name/namespaces/ns_id/enable"
on the target side was changed, the SPDK initiator should receive an
async event (type: SPDK_NVME_ASYNC_EVENT_TYPE_NOTICE, info:
SPDK_NVME_ASYNC_EVENT_NS_ATTR_CHANGED).
But actually not.
Since for SPDK, when target sent the non-first event, the condition
"nvmet_aen_bit_disabled(ctrl, NVME_AEN_BIT_NS_ATTR)" that prevents
target from sending event was matched.
This commit fix this issue by issuing a get_log_page cmd for each async
event received, just as the kernel initiator does.

Fixes #1825.

Signed-off-by: tyler.sun <tyler.sun@dell.com>
Change-Id: I2973470a81893456ca12e86ac390ea1de0eed62c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7107
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: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
2021-06-08 08:28:06 +00:00
Jim Harris
3a0f624499 nvmf: add USDT probes for subsystem state changes
Also add scripts/bpf/nvmf.bt to enable and log these
probes.

This patch also adds a script that can generate
a bpftrace script snippet with string maps for
needed enumerations (currently nvmf_tgt_state and
spdk_nvmf_subsystem_state).  This allows us to
dynamically generate this from the source code, and
can be extended for other enums we may want to
add in the future.

Thanks to Michal Berger for converting my original
gen_enums.py script into gen_enums.sh!

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Signed-off-by: Michal Berger <michalx.berger@intel.com>
Change-Id: Iff34a6218aef40055ac14932eea5fc00e1c8bcf5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7194
Community-CI: Mellanox Build Bot
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: Konrad Sztyber <konrad.sztyber@intel.com>
2021-06-08 08:27:15 +00:00
Jim Harris
3a7a66e505 nvme: add wmb before writing shadow register
Fixes issue #1828.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I02ba897b6c10f4ec870f46825f09e00f0c7c0235
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7905
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2021-06-07 23:34:30 +00:00
Jim Harris
c754043946 Revert "thread: speed up io_device lookup by using rbtree"
This reverts commit 2246a93718.

We are seeing a lot of failure on io_device lookup in the test
pool.  These only showed up after this patch was merged and sees
the most likely culprit.  Reverting this patch for now while we
continue debug.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I2ab098319dfae3a5356eb4fe0dbf9f4af2d2eea5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8199
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.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>
2021-06-07 16:17:33 +00:00
Jiewei Ke
2246a93718 thread: speed up io_device lookup by using rbtree
Use the macros for red black tree provided by Free BSD to speed up
io_device lookup.

Signed-off-by: Jiewei Ke <jiewei@smartx.com>
Change-Id: Ib3bd382bbeb610503194e7d7bfd569f60a0d0121
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7894
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-06-04 22:46:40 +00:00
Changpeng Liu
45b5676738 nvmf/vfio-user: rename vfio_user_stop_ctrlr() to vfio_user_destroy_ctrlr()
The original function will disconnect queue pairs first and then free
controller memory finally, so rename it to vfio_user_destroy_ctrlr().

Change-Id: Idc235e4186bd4164be712fc9d4cda4991efc6248
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7624
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: <dongx.yi@intel.com>
2021-06-04 22:14:23 +00:00
Changpeng Liu
0cc8bcf508 nvmf/vfio-user: rename destroy_ctrlr() to free_ctrlr()
The coming destroy_ctrlr() function will disconnect
queue pairs and free controller at last, so here
rename the original destroy_ctrlr() to free_ctrlr().

Change-Id: I527b2742142d60b0383be5a12391c77dd50d47a7
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7623
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: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: <dongx.yi@intel.com>
2021-06-04 22:14:23 +00:00
Changpeng Liu
3d4b9ba0d4 nvmf/vfio-user: rename destroy_qp to free_qp
The original destroy_qp() only release the queue pair
related memory, and free_qp() will be called inside
destroy_ctrlr() function, so also remove one duplicated
line here.

Change-Id: I2a06a6704b514361685068acda4e65ed5d502f0d
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7622
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: <dongx.yi@intel.com>
2021-06-04 22:14:23 +00:00
Changpeng Liu
12de3673f0 nvmf/vfio-user: remove controller ready field
The controller concept in NVMf is like a session, for any
new connection in nvmf_vfio_user_accept() with the endpoint,
we treat it as a new controller, we don't need the `ready`
field in controller to indicate the connection state, we
need the connection state in endpoint, so here just use
endpoint->ctrlr point to indicate the socket connection
is valid or not.

Change-Id: I588dbba7973cb61a1d79d81324a43e052f7dafb0
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7621
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-06-04 22:14:23 +00:00
matthewb
6127461c93 lib/bdev: Added iov to spdk_bdev_zcopy_start
Adding iov to the spdk_bdev_zcopy_start function enable spdk_bdev_zcopy_start to
be used by transport layers as the iov is owned by the transport command

Signed-off-by: matthewb <matthew.burbridge@hpe.com>
Change-Id: I6d2be7f49566048bf25b7711ada8d2fb49fea6ee
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6816
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: Michael Haeuptle <michaelhaeuptle@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-06-04 20:10:07 +00:00
matthewb
51d7f9b13f lib/bdev: Removed ZCOPY emulation
ZCOPY emulation is not required. Modules can check if the bdev module
supports ZCOPY.  If not supported the module uses the existing
READ and WRITE operations.

Signed-off-by: matthewb <matthew.burbridge@hpe.com>
Change-Id: Idac0a4d27a79a6c7e567c420e15637e826c347c8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6815
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Michael Haeuptle <michaelhaeuptle@gmail.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Community-CI: Mellanox Build Bot
2021-06-04 20:10:07 +00:00
Jim Harris
b333f00627 nvme: save last transport_failure_reason in transport
If a reconnect fails, we restore the original
transport_failure_reason after we're done with
the failed reconnect.  Save the original reason
in the qpair itself rather than a local variable,
to facilitate upcoming changes where connect will
be asynchronous.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I20ff43fc687a379aa5c930e17cf3ff8d730320be
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8116
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-06-04 20:03:25 +00:00
Jim Harris
43b4883217 nvme: don't reinit qpairs for vfio-user during reset
Previously we were only checking trtype==PCIE to
determine whether a controller was fabrics.  This
skipped the vfio-user case.  So use the new
spdk_nvme_transport_id_is_fabrics() API instead.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I81f26853f44b1c47522ce6354e5aa4a905796bd0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8089
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-06-04 20:03:25 +00:00
yupeng
b832f99f88 nvmf: Add nvmf_set_crdt RPC
Set the three CRDT values at SPDK_RPC_STARTUP time.

Signed-off-by: Peng Yu <yupeng0921@gmail.com>
Change-Id: I2fb4c4a3e367a4888cfec4658e6bf6899c7ae1f4
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8007
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: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-06-04 20:02:42 +00:00
Ziye Yang
26004a40fb accel_engine: Use the vector based crc32c function.
Purpose: Use the new function in order to reduce duplicated code.

Change-Id: Ie848c7586575b3f0bb617d7e767cf459b43d4783
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8174
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: Ben Walker <benjamin.walker@intel.com>
2021-06-04 20:01:20 +00:00
Curt Bruns
522faef3e2 nvme: add cuse support for RESCAN IOCTL
Nvme-cli submits a RESCAN IOCTL after a format command to
update any information that may have changed during the
format, such as LBA Format.  This patch adds support
for RESCAN by executing nvme_ctrlr_update_namespaces to
update the controller information.

Fixes: #1964

Change-Id: I9f03e00a7f39339947ff02390f69ce806e1cfa0e
Signed-off-by: Curt Bruns <curt.e.bruns@gmail.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8146
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>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-06-04 19:58:41 +00:00
Tomasz Zawadzki
47a6578ee5 lib/event: calculate last_stats in event framework
Rather than to rely on schedulers to access and modify
last_stats values over multiple scheduling periods, move that
operation to event framework.

Providing this to the schedulers in generic manner is better
than enforcing that each scheduler has to keep track of this
data on their own.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Icaf3b4af80d86fafaddf328fd230db9743d21ab5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7971
Tested-by: SPDK CI Jenkins <sys_sgci@intel.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: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
2021-06-04 19:58:24 +00:00
Changpeng Liu
a7f526ba1a nvme/opal: check Security Protocol 00h before sending TCG SP
A Security Receive command with the Security Protocol field cleared to
00h shall return information about the security protocols supported by
the controller, so we can check the TCG security protocol is supported
or not before sending it.

Fix issue #1961.

Change-Id: Id061defe45db981b276e2794fd0b59f8db70b7f4
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8083
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Niklas Cassel <niklas.cassel@wdc.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-06-04 19:57:36 +00:00
paul luse
f31ad9b6f1 lib/idxd: add cache control flag to op codes that support it
Perf improvement, directs DSA to write to cache as opposed to
mem.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I0d6ba157af8f1b54f8aae3b8e54a6f7754e4a9de
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8169
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: Ben Walker <benjamin.walker@intel.com>
2021-06-04 15:48:22 +00:00
paul luse
2e8ae7294a lib/idxd: stop polling on the first uncompleted location
We already keep a list of outstanding completion locations to
poll and were previously polling all of them.  New ones are
added at the tail and we poll the oldest first from the head
so if we break when we find a slot that hasn't completed we
can get more work done while the HW finishes. This is a proven
performance improvement in limited testing.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: Icc15041605586f9a31435d447d253c381c00b1f8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8161
Reviewed-by: Ben Walker <benjamin.walker@intel.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>
Community-CI: Mellanox Build Bot
2021-06-04 15:48:22 +00:00
Changpeng Liu
c52e700774 libvfio-user: update submodule
1. struct pxdcap.per is changed to struct pxdcap.rer
Which matches the name in the nvme spec.
2. use new API return value.
3. update specification changes.

Signed-off-by: Swapnil Ingle <swapnil.ingle@nutanix.com>
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Change-Id: Ida421c4cffd1c65d550e83011ab123b321ea9dff
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8088
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-06-03 04:03:04 +00:00
Changpeng Liu
b70b16acba nvmf: don't use request internal variable after req_complete callback
vfio-user transport `req_complete` callback will zero the internal
NVMe command and response fields, the common NVMf library should
not use them after the callback, so here we use stack variables
to save them before the `req_complete` callback.

Fix issue #1965.

Change-Id: Iff2342b6095d9496cdf112d657a0a99ce1fb5d12
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8129
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-06-03 04:03:04 +00:00
Shuhei Matsumoto
a4f96d93a8 nvme: Initialize ANA state to optimized if ANA reporting is not supported
Initialize ANA state of each namespace to optimized regardless of
whether ANA is supported or not. This will simplify the code to get
the optimal I/O path because we do not have to care if the namespace
supports ANA.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I24dfe08674af398671de6528b884e9d82409eeae
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7890
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-06-02 11:37:58 +00:00
Ziye Yang
b85127cc6f util/crc32c: Add spdk_crc32c_iov_update.
Purpose: To support caculating crc32c for iovs.

Change-Id: I7ea6d8b71ea2cf6e8b81903439787870923b9bef
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8093
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2021-06-02 11:37:17 +00:00
Ziye Yang
ee8f6dd1d5 sgl.h: Add a new internal header file.
The purpose is to reduce the duplicated code in nvmf and iscsi
layer.

Change-Id: I7e96f0d5bb1ba4b81378addca3cdd929056384e9
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8132
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: <dongx.yi@intel.com>
2021-06-02 11:34:27 +00:00
Jim Harris
9f5e3c99f2 nvme: add spdk_nvme_ctrlr_is_fabrics()
We map the SPDK_NVME_TRANSPORT_* values directly to
the NVMe-oF trtype values.  Since PCIe isn't
Fabrics, we choose 256 which is outside of the
8-bit trtype range of values.

So we can just check if trtype >= 256 to determine
if the trid is for fabrics or not.  This is
preferable to checking PCIE || VFIOUSER in case
additional non-fabrics transport types are added
in the future.

I considered taking a trid as the parameter instead,
but went this route since it is consistent with
the existing spdk_nvme_ctrlr_is_discovery().

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ib62ff4d30549b2324486c81f2dce67f0f1741e9b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8077
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2021-06-01 07:43:12 +00:00
Jim Harris
d6f6ffd274 nvme: add NVME_CTRLR_STATE_CONNECT_ADMINQ
Connect the adminq as part of controller initialization
instead of controller construction.

We never actually 'connected' the adminq for
PCIe or vfio-user transports, since its a nop.
But their connect_qpair transport ops function
is also a nop for the adminq, so it's fine to
generically connect the adminq across all transports.

Note that we cannot read registers (cc or csts)
during controller initialization now until after
the adminq has been connected since reading fabrics
registers depends on a connected adminq.  This gets
special cased for now, but eventually reading
cc and csts will need to be part of the state machine
itself to make it asynchronous.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia5566d7c549d78d24b94ea253df51e697da6237f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8079
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2021-06-01 07:43:12 +00:00
Ziye Yang
252430a053 nvme_tcp: Correctly handle the data digest err
According to NVMe-oF 1.1 spec, it is not a fatal error.
So according to Figure 126 in NVMe Base specification,
we should return "Transient Transport Error".

Change-Id: I601304ae2bb24508882fb1ec8c7e53ec587ab515
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7795
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2021-05-31 07:15:16 +00:00
Jim Harris
a31e319fbb nvme: init discovery ctrlr using nvme_ctrlr_process_init
This ensures the discovery ctrlr initialization is
done the same as normal ctrlrs.  This will be
critical as we make the driver fully asynchronous.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I33c4fd7c82d241c30e7adb89abe79b8088c8776a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8090
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-28 08:14:06 +00:00
Jim Harris
f5ba8a5ef5 nvme: add NVME_CTRLR_STATE_READ_CAP
Read CAP (Capabilities) register as part of controller
initialization instead of controller construction.

For now, still read CAP in the pcie and vfio-user
controller construction, since they need the
drstd (doorbell stride) to construct the admin
queue.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I000fe880f2ec0d6de1d565c883d7ea0ae1ac2c81
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8078
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-28 08:14:06 +00:00
Jim Harris
df01076f70 nvme: add NVME_CTRLR_STATE_READ_VS
Read VS (Version) register as part of controller
initialization instead of controller construction.

This prepares for upcoming changes to make
controller attach fully asynchronous.  Since reading
fabrics registers is an asynchronous operation, it
will be easier to read the VS register as part of
controller initialization which operates as an
asynchronous state machine.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I771386dbdf5902633e0d9f91b3b20be98f26fdc3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8076
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-28 08:14:06 +00:00
Jim Harris
a12cf26cdd nvme: add NVME_CTRLR_STATE_CHECK_EN
We're going to be adding some new states (READ_CAP
and READ_VS) in future patches, that we want to
come before the current "INIT" state.

So we will simply make "INIT" have the same
value as this new NVME_CTRLR_STATE_CHECK_EN state
for now.  That means existing code won't have to
change later once we add new states that come
before CHECK_EN.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I07ca92e28ab1cd8d838cdef5c3ff36ba80a224bf
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8075
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-28 08:14:06 +00:00
Tomasz Zawadzki
3db73426f2 lib/event: update current_stats at the time of thread reschedule
_reactor_schedule_thread() zeroes out the lw_thread on move.

To properly calculate thread stats since the move,
save them right after reschedule.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I44cc3b5907adda35b3117c2dd7268dc813d59853
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7919
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-28 08:13:48 +00:00
Tomasz Zawadzki
e0d448e7f6 lib/event: remove snapshot_stats
spdk_thread keeps track of tsc from its whole lifetime,
those can be requested with spdk_thread_get_stats() at any time.

spdk_lw_thread uses stats from above and keeps track of two points in time:
- current_stats reflecting stats at the time of gather_metrics stage
- last_stats reflecting stats from previous gather_metrics stage

1)
Before this patch current_stats were duplicated in snapshot_stats.
There is no need for that so now they are removed.

2)
Removed _spdk_lw_thread_get_current_stats() since it would be copying
current_stats to current_stats, thus not perform any action.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I5e5d4039cd0f7cc10ba150a3d915b90ec96589d7
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7842
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-28 08:13:48 +00:00
Tomasz Zawadzki
0598e484d7 lib/event: set thread for each event executed
Whenever an event executes, it might change the currently
set thread or reset it to NULL.
To prevent it from affecting other events, set the current
thread each time an event executes.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I6f1e7f8b7acab25353b4782058e87a9e01aab2c8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8045
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>
2021-05-28 08:13:48 +00:00
Tomasz Zawadzki
a4bf3e1099 scheduler_dynamic: exit early from _get_thread_load()
_get_thread_load() is function used to determine
the load of a thread based on relation of busy/idle tsc
from previous scheduling period.

In order to avoid division by 0 calculating the percentage,
we can simply exit early determining that thread was not
doing any work.

Having this check here will make sure that no matter
the changes in event framework, scheduler dynamic will work.

Removed the place that updated last_stats if they weren't
yet updated at least once (first scheduling period iteration).
In this case after change to _get_thread_load() will be the same,
as only the latest iteration will be used to calculate thread load.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I75f0f12f024675f2473a26e30596d6eb28093d46
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7917
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
2021-05-28 08:13:48 +00:00
Changpeng Liu
cfff04aea2 nvme/cuse: add subsystem reset IOCTL support
Change-Id: Id85989d0ae1246cb60fff4a41e5847a2d84e271f
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8084
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: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2021-05-28 07:20:23 +00:00
Rui Chang
080118cd83 lib/nvmf: Add current qpair count to poll group statistic
In struct spdk_nvmf_poll_group_stat, there are statistics of cumulative IO and
admin queue pair counts. But current qpair counts are not reflected. Use
this patch to add current admin and io qpair counts for a poll group.

Signed-off-by: Rui Chang <rui.chang@arm.com>
Change-Id: I7d40aed8b3fb09f9d34e5b5232380d162b97882b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7969
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Eugene Kochetov <evgeniik@nvidia.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: GangCao <gang.cao@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-27 07:35:49 +00:00
Changpeng Liu
75a507736b vhost-blk: use device name for VIRTIO_BLK_T_GET_ID
Block device product name is same among same type
of the block devices, while Guest VM may use this
value to generate UUID, so here we change it to
block device name instead.

Change-Id: I58c5fb271a6a436c15520616c2065eee9c37300a
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7996
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-27 07:35:37 +00:00
sunshihao520
c1a41845cf lib/bdev: Change spdk_bdev_read_blocks_with_md arg offset definiton from int64_t to uint64_t.
Signed-off-by: sunshihao520 <sunshihao@huawei.com>
Change-Id: I25c3ae4f12d0ab4e2f3811e36028a04a1f32d862
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7869
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-27 07:35:17 +00:00
John Levon
940f06490c vfio-user: update for libvfio-user region offset API change
Now, vfu_setup_region() must specify the region fd offset (which is always zero
in our case).

Signed-off-by: John Levon <john.levon@nutanix.com>
Change-Id: I10795d848a4c73ee9e1e78ea63776074401c4b17
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8022
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-27 01:33:31 +00:00
Jiewei Ke
eabe783cc4 bdev: speed up bdev name lookup by using rbtree
Use the macros for red black tree provided by Free BSD to speed up bdev
name lookup in spdk_bdev_get_by_name().

In the bdev_multi_allocation test, we can get 3x ~ 5x speed up when
creating multiple bdevs for various bdev nums.

Signed-off-by: Jiewei Ke <jiewei@smartx.com>
Change-Id: I49a2fbcccf06d4c36cbd445ce59e0b0dd4ada31d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7837
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:22:06 +00:00
Jiewei Ke
1a22f9b6ac bdev: clean up spdk_bdev_register()
Clean up spdk_bdev_register() to facilitate the upcoming patch which
uses RB tree to speed up bdev name lookup.

* move TAILQ_INSERT from bdev_start() into bdev_init();
* rename bdev_init() to bdev_register() and rename bdev_start_finished()
  to bdev_register_finished();
* inline bdev_start() into spdk_bdev_register().

Signed-off-by: Jiewei Ke <jiewei@smartx.com>
Change-Id: Idbfc800472bc8c6f9b615046e082772e9f6026e3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8043
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:22:06 +00:00
Jiewei Ke
d07c581b7d nvmf: nvmf_get_ana_log_page should use iov to store the log page
nvmf_get_ana_log_page used req->data to store the log page result.
While the req->data only contains the first iov, if req->iovcnt is
larger than 1, the req->data may not hold the complete log page; and
even worse, the log page result may be written to invalid address and
cause memory corruption.

The following patch will fix the same issue for other commands in
nvmf_ctrlr_get_log_page.

Fix #1946

Signed-off-by: Jiewei Ke <jiewei@smartx.com>
Change-Id: I495f3be05c82be5cd53609772c655c8924b9179f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7923
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: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:20:30 +00:00
Ziye Yang
2250abaeca nvme/tcp: Raname send_pdu to pdu in tcp_req.
Since we will reuse send_pdu for other purpose in the next
patch.

Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: Iee5166131b70a25bc13aaa847bfc9066231f31a9
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8028
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:20:15 +00:00
Ziye Yang
9776b89444 nvme/tcp: Fix the bug when doing offloading.
For nvme/tcp connection, we use the synced manner
if the qpair is not fully connected. Thus without
the check, we will stuck here. And this patch
fixes this issue.

Change-Id: I72815bf5b4c0b31c4866bc1b9034b0e42b81d3f1
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8025
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:20:15 +00:00
Ben Walker
dde419084b event: Move json config loading to init
Loading subsystems and restoring state from a JSON config file is useful
outside of the SPDK application framework, so move it to lib/init.

Change-Id: I7dd3ceace2e7b1b28eef83c91ce6a4eedc85740e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6645
Reviewed-by: Tom Nabarro <tom.nabarro@outlook.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:13:34 +00:00
Ben Walker
9f62796127 event: Move RPC server initialization to init
I'm not sure whether this should go into lib/init or to lib/rpc
directly, but I've chosen lib/init for now.

This is to support applications that want to run the SPDK JSON
RPC server, but aren't using the SPDK application framework.

Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: I79ca39aa0ca6e1a3a6905b0bf73e6cc99b086e55
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6644
Reviewed-by: Tom Nabarro <tom.nabarro@outlook.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:13:34 +00:00
Ben Walker
dc9374dd2f init: Declare a formal public API
The functions to initialize the SPDK subsystems or tear them down
was previously an internal-only API. Make it public for use by
applications that aren't leverage SPDK's application framework.

Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: I2ebfd020e6fa4c1947fa1c1a2ac509ce9b0242f8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6643
Reviewed-by: Tom Nabarro <tom.nabarro@outlook.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:13:34 +00:00
Ben Walker
1a8467b02c init: Make some functions entirely private
There were some functions in the internal header that can be entirely
private to the init library. Move them over.

Also, remove the support for including the header from a C++ file
because these headers are internal to SPDK which is pure C.

Change-Id: Ic4323b2b8664e70106a57b3ca8acbc7c2efe621d
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6642
Reviewed-by: Tom Nabarro <tom.nabarro@outlook.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-26 09:13:34 +00:00
Ziye Yang
00b0dc6624 nvme/tcp: Do not offload header crc32c calculation if header digest is enabled.
The header size is very small, which does not have too much value to
offload such calculation by hardware.

Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: Iaa82f39312df7eef3282325a33677ea41ab735ab
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8011
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-25 07:12:43 +00:00
Tomasz Zawadzki
42ad9d0379 lib/blob: resize open_blobids on bs load
open_blobids holds bit array of currently open blobs,
this is a way for quicker determination than iterating
over all blobs. See patch introducing it:
(30ee8137)blob: Add a bitmask for quickly checking which blobs are open

That patch added resizes of this bit array to bs init
and bs recovery path (not shut down cleanly).

But that patch skipped over bs load from a clean shutdown.
This resulted in blob open having multiple blob pointers that
target the same blob id.

Fixes #1937

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I3c42a63d168d1f5b013b449f010c5b207936045b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7998
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Community-CI: Mellanox Build Bot
2021-05-25 07:12:07 +00:00
Tomasz Zawadzki
e65310624b lib/event: accumulate idle_tsc when reactor has no threads
Before this patch idle_tsc was sum of all idle tsc of all
threads running on a reactor.

There are cases when no threads are present on the reactor,
and _reactor_run() spins doing nothing.

To give more accurate representation of the reactors state,
the idle_tsc now adds time spent doing idle spinning.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: If797b2a03507d17b07367d56d5f6c40cefbbbd49
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7900
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-25 07:11:53 +00:00
Tomasz Zawadzki
a5ad0f8051 lib/event: update reactor tsc_last going poll mode
Disabling interrupt mode on reactor now updates the tsc_last
to current time. So any further tsc caulations will not account
for the time when reactor was in interrupt mode.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I56fb8a738eea60ee5de3b49d586f7cb228b54510
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7901
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-25 07:11:53 +00:00
Tomasz Zawadzki
7438c388a0 lib/event: fix tsc_last for first thread on reactor
tsc_last value is used to update thread stats
during _reactor_run(). See:
spdk_thread_poll(thread, 0, reactor->tsc_last);

If no threads were present on the reactor,
this value got outdated and resulted in
adding time reactor spent with no threads to
stats of the first thread placed on that reactor.

This patch fixes thread stats by making sure
that argument to spdk_thread_poll() is up to date.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I0c35fdba1b63b6ee19a5a2b34751090839cb2438
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7845
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
2021-05-25 07:11:53 +00:00
Ben Walker
af935f768e event: Shift subsystem initialization code to a separate library
This is useful for applications even if they elect not to use the SPDK
event framework.

This doesn't shift everything in one go - just the subsystem
initialization logic. Configuration file loading also needs to move
in a separate patch later.

Change-Id: Id419df1045442d416650ed90e5ee78adfdd623d7
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6641
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-24 10:12:50 +00:00
Ben Walker
15ae31fb0c nvmf: Rearrange spdk_nvmf_requset for cache line locality
Get all of the hot stuff to the first cache line.

* Shrink the xfer enum to one byte (it only has 3 values).
* Pull out the dif enabled flag form the dif structure so it
  can be access separately
* Rearrange the members

Change-Id: Id4a2fe90a49c055a4672642faac0028671ebfae9
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7827
Community-CI: Broadcom CI
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: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-24 10:10:19 +00:00
Tomasz Zawadzki
1180c390c1 lib/thread: fix iterating over paused pollers
(cd83ea4a)thread: Add SPDK internal APIs spdk_thread_get_first/next_active/timed/paused_poller()
Patch above by mistake iterates over active_pollers list
for function that lists paused pollers.

Fixes #1947

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I1b69d942675f34f5f046ec46feacc8d81d89f015
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7952
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: <dongx.yi@intel.com>
2021-05-24 10:08:34 +00:00
Tomasz Zawadzki
ceaa0c7fa9 lib/blob: complete multiple persists
When blob persist starts, there can already be multiple
of such requests pending. It is possible to complete
a set of persists at once, if blob state after their
execution would be the same. This is the case when
persists are already pending when a particular persist
request is started.

This patch implements such mechanism by introducing
persists_to_complete queue, containing entries that
were previously queued up before starting the current
persist request. If there are any entries in this queue,
further requests are put into pending_persists.

When first request from persists_to_complete is persisted,
completions are issued for all requests on that queue at once.
If at that point there are any new entries on pending_persists,
all of them are put into persists_to_complete. Persist process is started
again with the first request from that queue.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I10063e55d6f821b1863de016d3148da6a719a422
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7643
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>
2021-05-24 10:08:00 +00:00
paul luse
eb5a111398 lib/idxd: move the crc destination * from the desc to comp record
Was using reserved field in CRC to store the final address of where
to put the result, this not legal.  Move to the completion record
and slightly re-arrange the struct to keep it at 96 bytes.

Refacorted the IO prep function so the caller can udpate both the
descriptor and completion records instead of continuing to add
parameters to the prep function for opcodes that need something
unique in the completion record.  This also allowed for a minor
fix where the prep function was returning NULL when vtophys failed
which would have indicated busy as opposed to failire.  Now we can
proprely fail that path.

fixes #1929

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: Ic23bc7b68bdd5757c30b7963880677f423368e20
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7735
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>
2021-05-22 18:42:35 +00:00
John Levon
c41508b7e2 libvfio-user: update submodule
Update libvfio-user to the current version, updating the client for the relevant
changes.

Signed-off-by: John Levon <john.levon@nutanix.com>
Change-Id: Ic64ace08ac0c7e9676f04f8d1f47a9c0388a2652
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7983
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Community-CI: Mellanox Build Bot
2021-05-21 03:01:21 +00:00
yupeng
5cc565999e nvmf: support NVME ACRE feature
Support ACRE (Advanced Command Retry Enable) feature. Currently set
all crdt to 0 and only select crdt[0] (crd=1) when the IO has any
error.

Signed-off-by: Peng Yu <yupeng0921@gmail.com>
Change-Id: If7bc30f91f5b2d0839002dead17188a4b3a52d5d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7885
Community-CI: Broadcom CI
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: Ben Walker <benjamin.walker@intel.com>
2021-05-20 15:18:52 +00:00
Alexey Marchuk
103f5c62f5 nvmf: Remove qpair from poll group when it is deleted
That prevents nvmf target from starting to destroy poll
groups prematurely

Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: I833f6198ef0e3083fdadf70dd3b62844c905aceb
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7881
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
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>
2021-05-20 15:18:29 +00:00
Evgeniy Kochetov
f0e248586e nvme/ctrlr: Always access namespaces via getter function
Signed-off-by: Evgeniy Kochetov <evgeniik@nvidia.com>
Change-Id: I16e327c0c0485057dc90a87cae316c4d6b62720d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6504
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-05-20 15:17:41 +00:00
Evgeniy Kochetov
65ff07719d nvme/ctrlr: Retrieve active NS list before NS construct
This patch changes the order of IDENTIFY_ACTIVE_NS and CONSTRUCT_NS
controller states. It is required to further improve memory management
for namespaces by allocating memory only for active ones.

Signed-off-by: Evgeniy Kochetov <evgeniik@nvidia.com>
Change-Id: Ie540442b1bd9e897afcbaa4319c139109dd0c515
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6503
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-05-20 15:17:41 +00:00
Evgeniy Kochetov
75ddfd6511 nvme/ctrlr: Handle namespaces construction failure
Signed-off-by: Evgeniy Kochetov <evgeniik@nvidia.com>
Change-Id: I3b14e2dce3bd4b2ce42e3946592cfde9d4300f28
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7361
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-05-20 15:17:41 +00:00
Evgeniy Kochetov
aebbce2520 nvme: Refactor active namespace list retrieval
Previous implementation allocated memory just once at the beginning of
active NS list retrieval procedure. It allocated memory for maximum
possible number of active namespaces, i.e. 'cdata.nn'.

This patch changes allocation logic. One page is allocated at the
beginning. If more is needed, reallocation is done with one more
page.

This patch also removes SPDK_MALLOC_DMA flag from allocation since we
don't do RDMA directly into this buffer.

Signed-off-by: Evgeniy Kochetov <evgeniik@nvidia.com>
Change-Id: Iaa80c4d70c54daaf71dcbf755c63a01a1d83b772
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6502
Community-CI: Broadcom CI
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>
2021-05-20 15:17:41 +00:00
Shuhei Matsumoto
4e9adb3bf9 thread: Replace TAILQ by Red Black tree for timed pollers
Use the macros for red black tree provided by Free BSD to manage
timed pollers efficiently.

Allow RB_INSERT() to insert elements with duplicated keys by changing
the compare function to return 1 if two keys are equal.

Check the return code of RB_INSERT() because this is the first use case
for RB tree macros in SPDK. We did the same for RB_REMOVE() by
adding another temporary variable but we remove it from this patch
because it is not so important compared with RB_INSERT().

When a timed poller is inserted, update the cache for the closest (leftmost)
timed poller only if the tree was empty before or the closest (leftmost)
timed poller was actually changed. We do not have to use RB_MIN()
because all duplicated entries are inserted on the right side.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ibe253ca8eecc10116548b5eedbcdba8fb961b88d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7722
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: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-20 15:13:27 +00:00
Shuhei Matsumoto
4eb96aeba0 thread: Inline poller_remove_timer() into thread_poll()
We already hold thehe next closest timed poller in tmp. Inlining
poller_remove_timer() into thread_poll() makes the cache update
more efficient.

After this patch, poller_remove_timer() is called only in a single case
and the case is compiled only on Linux. So add it inside of a temporary
block is much clearner. However it will be used by spdk_poller_reschedule()
in the end of this patch series. So keep
the current position.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I2e6858223713eed84f5d70b160da6122edae6d03
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7910
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: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-20 15:13:27 +00:00
Shuhei Matsumoto
83c197633f thread: Consolidate poller_remove_timer() calls into a single place
This enable us to optimize the cache update when RB tree is supported.

Call poller_remove_timer() after getting the next element because
as TAILQ_FOREACH_SAFE() and RB_FOREACH_SAFE() do, TAILQ_NEXT() may
not be valid after the current element is removed.

Previously, the patch had called poller_remove_timer() before getting
the next element. However, thanks to the nice testing, this bug was
found.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I18afb4412115dc1696cc568610cbe3dc618c2357
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7909
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: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-20 15:13:27 +00:00
Shuhei Matsumoto
3ca15e33e4 thread: Set poller to PAUSING even when it is in RUNNING
This change is a preparation to first dequeue the closest timed poller
always when it is expired. Previously the poller_remove_timer() calls
were not consistent and difficult to follow.

spdk_poller_pause() sets poller to PAUSING even when it in RUNNING
and move it to PAUSED after returning from its context.

If spdk_poller_pause() and spdk_poller_resume() are called while poller
runs, it is moved to WAITING. Hence thread_execute_poller() and
thread_execute_timed_poller() ignore such cases.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I29340613a2ec0c3529d0886f4d81c0a0fdf8745d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7908
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-20 15:13:27 +00:00
Jim Harris
f651e22a98 bdev: remove all references to spdk_bdev_open
This API was removed previously, so remove remaining
references in map file and unit tests.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Iba2f6a5f5ba590d3996dc133c8181083a33d7405
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7963
Community-CI: Mellanox Build Bot
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: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-20 15:02:59 +00:00
Alexey Marchuk
6263457ca3 nvmf: Remove spdk_nvmf_request_get_buffers_multi API
It is not used.

Change-Id: I238c5e2b2d557063f8443ca703551e02b35ba2c7
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7882
Community-CI: Broadcom CI
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>
2021-05-18 12:44:18 +00:00
Alexey Marchuk
019a5361a5 nvmf/rdma: Use contig buffers for multi SGL payload
Currently we allocate buffers perf each SGL descriptor.
That can lead to a problem when we use NVME bdev with
PRP controller and length of the 1st SGL descriptor is
not multiple of block size, i.e. the initiator may send
PRP1 (which is SGL[0]) which end address is page aligned
while start address is not aligned. This is allowed by
the spec. But when we read such a data to a local buffer,
start of the buffer is page aligned when its end is not.
That violates PRP requirements and we can't handle such
request. However if we use contig buffer to write both
PRP1 and PRP2 (SGL[0] and SGL[1]) then we won't meet
this problem.

Some existing unit tests were updated, 1 new was added.

Fixes github issue #1853

Change-Id: Ib2d56112b7b25e235d17bbc6df8dce4dc556e12d
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7259
Community-CI: Broadcom CI
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>
2021-05-18 12:44:18 +00:00
Richael Zhuang
c07a6a949b event/reactor: fix bug when showing core state
It should print "intr" when target->in_interrupt==1.

Change-Id: I74479da1dc66d937899fdf5c619cdf1a09951918
Signed-off-by: Richael Zhuang <richael.zhuang@arm.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7857
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>
2021-05-18 12:44:02 +00:00
Jim Harris
199fa6015d util: add zipf random number generator
zipf is a power law probability distribution. When
applied to performance testing of block devices, it
will select blocks over the full range of LBAs, but
will more frequently select lower-numbered LBAs.
The theta parameter governs the distribution - higher
values of theta will concentrate the distribution on
a smaller number of LBAs.

Note that fio supports zipf, so adding it to SPDK
will enable our perf tools (bdevperf, nvme-perf) to
provide similar functionality.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I7df129c9d61996a2070188c6cd9f1fde631ac208
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7779
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-17 11:29:52 +00:00
Jim Harris
da9766336e nvmf: delay remove subsystem cb until no qpairs remain
We cannot solely rely on the qpair_ctx->count reaching
0, because qpairs that are in process of being
disconnected will immediately invoke the qpair
disconnect cb.

Instead, we need to wait until the poll group
no longer has any qpairs remaining on the subsystem.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I977747d367d14a4bf60f66a1147b3d75679e5179
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7870
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: <dongx.yi@intel.com>
2021-05-17 09:52:37 +00:00
Shuhei Matsumoto
4748ebef40 thread: Cache the closest timed poller into thread
When we introduce RB tree, getting the closest timed poller is not
O(1) but O(log N). To mitigate such delay, cache the closest timed
poller into thread, and update the cache when its content is changed.

Add unit test cases for this change. They will also clarify the current
behavior of spdk_poller_unregister() and spdk_poller_pause() for
timed pollers.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ibb98a54c261859a3210034038d3953e5c93ef8aa
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7720
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Community-CI: Mellanox Build Bot
2021-05-14 10:08:02 +00:00
Wojciech Malikowski
180d281c7f lib/ftl: Drain retry queue before shutdown
Requests that still reside on retry queue should be
submitted to disk before shutdown.

Signed-off-by: Wojciech Malikowski <wojciech.malikowski@intel.com>
Change-Id: Id2d020fcaef6443d01cfd8628686e9b0f34a1cfa
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6771
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-14 10:05:34 +00:00
Alexey Marchuk
4642d7b264 nvmf/rdma: Embed fill_wr_sge into fill_wr_sgl
This patch reduces admount of changes in the next patch,
no functional changes added.
The next patch will add usage of contig IO buffers for
multi SGL payload. To support it we need to pass an
offset to fill_wr_sgl function. Also in the current
version we assume that for 1 iteration we fill 1 IO
buffer, the next patch will change it and we'll need
to swtich to the next IO buffer in special case. That
can't be done easily if we use fill_wr_sge function

Change-Id: Iee8209634637697f700f8fa9fe61ead156b6d622
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7258
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-14 10:03:56 +00:00
Ben Walker
64a9432c00 nvmf/tcp: Don't attempt to offload header digest calculations
The header is small enough that it likely won't ever make sense
to offload the digest computation.

Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: Ib6baa201a76d769d978f498f5c65985d5ab06ffd
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7766
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-14 10:03:05 +00:00
Krishna Kanth Reddy
7a12f4812d lib/nvme: Add support for Simple Copy Command
Signed-off-by: Krishna Kanth Reddy <krish.reddy@samsung.com>
Change-Id: I7aae5abca6ba5621e99fd604d146d84401eb45cd
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7694
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
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>
2021-05-14 09:59:54 +00:00
Changpeng Liu
6b372f419d nvmf: fix heap corruption of Reservation Report command
We should use `ns->registrants` to count the number of registered
controllers(REGCTL) for Reservation Report command, as `subsystem->ctrlrs`
only list current active sessions(controllers).

Also use the output data buffer directly so that we don't need to calloc/free
during the process of Reservation Report command.

Fix #1928

Change-Id: I650224b751a08416208b8a504b82debff31e92fd
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7822
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: wanghailiang <hailiangx.e.wang@intel.com>
2021-05-13 10:10:42 +00:00
Shuhei Matsumoto
54bc83dcc7 thread: Make struct spdk_thread and enum spdk_thread_state private in thread.c
Move the definition of struct spdk_thread and enum spdk_thread_state
from include/spdk_internal/thread.h to lib/thread/thread.c.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Iab8cec9776e76668ebfb3c75064c316d10607421
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7802
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
bd2fae2f0e thread: Change remaining direct accesses to io_channel outside lib/thread to helper functions
This is the same effort as spdk_poller.

The following patches will move the definition of struct spdk_thread and
enum spdk_thread_state from include/spdk_internal/thread.h to
lib/thread/thread.c.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I7f7bdfdd7a7b1b834d16d79638a4fd2d63e9daf6
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7800
Community-CI: Mellanox Build Bot
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: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
5fdb2b761b thread: Make struct spdk_poller and enum spdk_poller_state private in thread.c
Move the definition of struct spdk_poller and enum spdk_poller_state
from include/spdk_internal/thread.h to lib/thread/thread.c.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I4480a0b7a2a94ef97ff3185e458221e8c473665d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7799
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
1aec9334d9 thread: Change direct accesses to poller outside lib/thread to helper functions
Most accesses to the struct spdk_poller outside lib/thread have been
done via functions but a few direct accesses remain.

Change these to indirect accesses by addinng a few helper functions
as SPDK internal APIs.

Add spdk_poller_get_name() to get the name of the poller.

Remove spdk_poller_state_str() and add spdk_poller_get_state_str().
Exposing enum spdk_poller_state outside lib/thread is not really
necessary.

This removal requires us to update major SO version.

Add spdk_poller_get_period_ticks() to get the period ticks of the
poller.

Add struct spdk_poller_stats and spdk_poller_get_stats() to get
the stats of the poller.

The next patch will move the definition of struct spdk_poller and
enum spdk_poller_state from include/spdk_internal/thread.h to
lib/thread/thread.c.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Id597dae074a15fcd8af09fd9d416a22ce2f403c3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7798
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
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: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
9ff6238b88 event/scheduler: Include not spdk_internal/thread.h but spdk/thread.h
It is better if the internal of poller or thread is not accessed
outside lib/thread.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I80d488a111fd9a67a0da32d1e63695ce5a6bcb4c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7776
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
f5313b5501 thread: Factor out removing timed poller from list into a helper function
The following patches will update the cache to the closest timed
poller when removing it from the list. To do it easier, factor out
the operation to remove a timed poller from timed_pollers list into
a helper function.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I25016d86117b240a2651d1f06e23bea0342211f1
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7719
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
c204c3d786 thread: Change TAILQ_FOREACH_SAFE to TAILQ_FIRST() and _NEXT() for timed poller
When we introduce red black tree for timed pollers, we will not use
RB_FOREACH_SAFE() but cache the leftmost (smallest) node and iterate
from it via RB_NEXT() instead.

As another preparation, separate TAILQ_FOREACH_SAFE() into TAILQ_FIRST()
and TAILQ_NEXT().

The next patch will cache the first element to thread and refer it
first.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ie03c387b5b3a055c668e7b439a5eb05ed77eaa81
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7718
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
aca41b43a3 thread: Factor out convert period from usec to ticks into a helper function
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I1dc0a202aab61a42511d3afac6b6533cfe6d1ade
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7666
Community-CI: Mellanox Build Bot
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: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
42ad32da60 thread: Change if-else_if to switch-case blocks of spdk_poller_pause/resume()
It will be sufficiently reasonable to check if the caller thread is valid
even if spdk_poller_pause() or spdk_poller_resume() does nothing.
Besides, let's write all possibles states explicitly in switch - cases.

This refactoring clarifies the logic and makes the following patches easier.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I1163eff388fe741d6b6924f474a82b1aa7d18acb
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7665
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
3f45ed2467 thread: Change if-else_if to switch-case blocks of thread_execute_poller()
Change if - else if - else blocks of thread_execute_poller() and
thread_execute_timed_poller() to switch - cases blocks and specify
possible states explicitly in these switch - cases blocks.

The code will be simpler and clarified, and then the following patches
will be easier.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I2e283894f5d69e1bd67466ae070c5b8bb9014616
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7664
Community-CI: Mellanox Build Bot
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: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
54215a1e9b thread: Factor out executing poller or timed poller into helper functions
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I8792b698f5de85c24482d6781fc3d47afce0c414
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7663
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
84ec998976 thread: Check if timed poller is expired first in iteration of thread_poll()
There will be no issue even if time poller is unregistered or paused
after it is expired. The iteration is stopped anyway after the head poller
is found not to be expired.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I2b394b8b517930a6630dd31f59fcaea12eb80572
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7662
Community-CI: Mellanox Build Bot
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: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
cd83ea4a6d thread: Add SPDK internal APIs spdk_thread_get_first/next_active/timed/paused_poller()
The following patches will introduce red black tree to manage
timed pollers efficiently but it will be based on macros available only
in lib/thread/thread.c. Hence then it will be difficult to expose the
internal of timed pollers tree outside the file. On the other hand,
we do not want to include JSON into the file.

Hence add a few SPDK internal APIs to iterate pollers list transparently.

For spdk_thread_get_next_active/timed/pause_poller(), we omit the parameter
thread and get it internally from poller->thread even if the names include
the term "thread". This will be slightly cleaner.

Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I000801a2e4dc42fa79801a2fd6f2b06e1b769c88
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7717
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-13 10:03:44 +00:00
Shuhei Matsumoto
d5b7f3c580 thread: Add check if the correct thread called spdk_poller_pause/resume()
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ie2156a331bc2384a1cebe66fcddb90ef5aa1c380
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7661
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: <dongx.yi@intel.com>
2021-05-13 10:03:44 +00:00
Jonathan Teh
be6a01efb3 nvmf: Allow configurable controller ID range
Add {min,max}_cntlid to spdk_nvmf_subsystem, defaulting to 1 and
0xFFEF, respectively, and add nvmf_subsystem_set_cntlid_range() to
allow the controller range to be configured in the range [min_cntlid,
max_cntlid].

Also add {min,max}_cntlid to the nvmf_create_subsystem RPC to allow
the controller ID range to be specified when creating an nvmf
subsystem.

Signed-off-by: Jonathan Teh <jonathan.teh@mayadata.io>
Change-Id: I936db3bb0c9a38569063a6fd3c11df262dfad776
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7322
Community-CI: Broadcom CI
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>
2021-05-12 09:18:48 +00:00
Mao Jiang
3067ecaa0c nvme/ctrlr: fix namespaces indexed overflow risk
Change-Id: Ied1de194621a02f6eeaafc23ce5ff467760a71aa
Signed-off-by: Mao Jiang <maox.jiang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7772
Community-CI: Broadcom CI
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: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-12 09:17:38 +00:00
Changpeng Liu
55189bee6b libvfio-user: update submodule
This update will stop using `struct vfio_device_info` from
<linux/vfio.h>.

Fix issue #1922.

Change-Id: Ia7ad745db8d7ed8f5248ca13e3188ebd540b0e40
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7831
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-12 08:46:29 +00:00
Tomasz Zawadzki
a8fe601853 lib/blob: delay executing blob persist completions
This helps in next patch in series where multiple
completions will be executing.

UT is adjusted since one additional poll is required.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Id72377ddef91e40cdbc2bdea6f33c23309b0ca3d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7642
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-11 12:03:57 +00:00
Tomasz Zawadzki
958364b4a0 lib/blob: only verify cluster allocation when I/O is not frozen
During snapshot creation the original blob becomes
a thin provisioned blob that will only the diff of data after
snapshot creation.

Despite the comment in the UT the number of polls before issuing
blob write was hitting blob BEFORE it swapped with new one.

Issuing I/O during this period shall check for io freeze
before checking cluster allocation.

Otherwise bs_io_unit_is_allocated() hits assert for thin
provisioned blob. This is because cluster map of blob is
empty, but properties have not been updated yet.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I742e1a50b14d456ae1e6de13b5111caec3e8322c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7641
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-11 12:03:57 +00:00
Changpeng Liu
218d6ac8cc bdev: add bdev_unmap_split() support
Also for the purpose to avoid a burst of children unmap request,
we will submit at most up to 8 children request at a time for
a big UNMAP command.

Change-Id: Iaf0f18b07517e0a8f84dc04e8c93b95691a1a43c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7518
Community-CI: Broadcom CI
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: Ben Walker <benjamin.walker@intel.com>
2021-05-11 12:03:28 +00:00
Changpeng Liu
e6db8e02e5 bdev: add a helper function to process IO split submission
Both UNMAP and R/W will share the some logic to process the submission,
so combine them to a helper function first.

Change-Id: Ia4f234c6a58f078d3e9f88cacaf1510a17f07acc
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7606
Community-CI: Broadcom CI
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>
2021-05-11 12:03:28 +00:00
Swapnil Ingle
2421e99849 scsi: Cache physical block size locally
Addressing review comment from 6cebe9d0.
Minor optimisation to cache result of spdk_u32log2() into local variable
instead of calling it multiple times.

Signed-off-by: Swapnil Ingle <swapnil.ingle@nutanix.com>
Change-Id: I2fd6afd1e3ee461662de3f9d278958664224e106
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7806
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: sunshihao <sunshihao@huawei.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
2021-05-11 11:59:31 +00:00
Jim Harris
53447f060f nvmf: remove a bdev_module.h #include
ctrlr_discovery.c doesn't need this #include.
Including it causes bdev_module.h types to be
emitted to the debug symbols at least with some
compilers, which can result in unwanted abidiff
errors.

The unit tests do need it, so just include it
there instead.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Iad28f9778ce08b11b52325658583ae9032295f3a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7813
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Michal Berger <michalx.berger@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-11 11:57:01 +00:00
Jim Harris
8a12e2e715 scsi: don't include bdev_module.h
The library itself doesn't need it. The unit tests
do need it, so just include it there.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9aefd303ae12928d45141029436509f185105bd3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7812
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Michal Berger <michalx.berger@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-11 11:57:01 +00:00
Robert Baldyga
d20f41881b spdk_zmalloc: Remove unnecessary memset()
Normally, unless RTE_MALLOC_DEBUG is set, DPDK zeroes memory in rte_free().
If RTE_MALLOC_DEBUG rte_free() fills memory with poison pattern, but then
(and only then) the memory is zeroed in rte_zmalloc_socket(). Relying on
this behavior allows to avoid unnecessary memset() in spdk_zmalloc() path.

Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
Change-Id: If3efa4dd22f1568949c3fb529b604bd597ceb32f
Signed-off-by: Rafal Stefanowski <rafal.stefanowski@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6975
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
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: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2021-05-11 11:49:06 +00:00
Ziye Yang
c94648ff91 lib/accel: Add the missed seed assignment for chain crc32c support.
We did not find this issue, because we always use seed = 0 to test.
Definitely, we need to assign the seed value.

Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: I6bec1b6d61480cdd7c9e27578dcaf5de2f65cf44
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7716
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: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2021-05-10 11:14:37 +00:00
Jonas Pfefferle
19faf91293 nvmf: ctrlr_create - fix missing free qpair_mask on error
This patch fixes missing free of qpair_mask when a listener error
occurs in ctrlr_create.

Signed-off-by: Jonas Pfefferle <pepperjo@japf.ch>
Change-Id: I09162b86d8ac73bf9fc2006a08dcc0a955f222b3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7818
Community-CI: Broadcom CI
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>
2021-05-10 11:14:24 +00:00
paul luse
e8fc521f7d lib/idxd: set error status on any completion error.
Not just one with extra available info.  Also remove the extra
read of the error register, not required.

fixes: #1927

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I28badb45d8cc8d16b72f7019bd2a2044998fc402
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7729
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-by: Krzysztof Karas <krzysztof.karas@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
2021-05-10 11:13:26 +00:00
Curt Bruns
b89ad84532 nvme: add support for NVME_IOCTL_IO_CMD for cuse
Nvme-cli uses NVME_IOCTL_IO_CMDs for "io-passthru"
commands to cuse devices.  This patch adds support
for that IOCTL.

Signed-off-by: Curt Bruns <curt.e.bruns@gmail.com>
Change-Id: I20e0ac91ba08fce91bc5da1f4a1e454058cdd1e7
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7741
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-10 11:13:09 +00:00
Curt Bruns
f3edd7a359 nvme: rename "admin" to "passthru" for cuse
The nvme cuse IOCTLs are actually creating passthru commands
that can be either IO passthru commands or admin commands.
Renaming the routines to correctly reflect that should limit
the confusion when reading the code.  Passthru commands that
are admin commands will go to the spdk_nvme_ctrlr_cmd_admin_raw
interface and passthru commands that are IO will be sent to the
spdk_nvme_ctrlr_cmd_io_raw interface.

Signed-off-by: Curt Bruns <curt.e.bruns@gmail.com>
Change-Id: I8d427fe8b5f503fdb2d193236c77d410d5b13886
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7740
Community-CI: Broadcom CI
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: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-10 11:13:09 +00:00
Curt Bruns
4c19332b0c nvme: add cuse support for BLKSSZGET IOCTL
Nvme-cli uses BLKSSZGET so support needs to be added for
nvme cuse devices.

Signed-off-by: Curt Bruns <curt.e.bruns@gmail.com>
Change-Id: Ic8316713b2d017c8ff32a225efff6bcb95842799
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7708
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
2021-05-10 11:13:09 +00:00
Curt Bruns
0ba168fe39 nvme: add SPDK_DEBUGLOG support for nvme_cuse
It is useful to have debug log information in the nvme_cuse
path when debugging IOCTls and flows.

Signed-off-by: Curt Bruns <curt.e.bruns@gmail.com>
Change-Id: Ifef1bb82c96438e2fcbb9ad2fafe3f3eb66bed51
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7707
Community-CI: Broadcom CI
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: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-10 11:13:09 +00:00
Ziye Yang
bcbccf8bb5 lib/nvme_tcp: Refactor the code to generate _nvme_tcp_pdu_payload_handle
The purpose is to prepared for implement the async crc32 caculation
in the future patch.

Change-Id: Ia75f28154c49f08b527d48c63b9da79a6bdfede8
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7794
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-10 11:12:57 +00:00
Ziye Yang
bc35dab91c lib/nvmf: Refactor the code to generate _nvmf_tcp_pdu_payload_handle
The purpose is to prepared for implement the async crc32 caculation
in the future patch.

Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: I49f84ea1966f0acdd6f5aeb7192896f91fd16dee
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7793
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
2021-05-10 11:12:57 +00:00
Ziye Yang
4e0ca20a8a idxd: Do not present pci device info in accel_engine_idxd.c module
Purpose: We will also support the kernel idxd driver, so we do not
need export this feature in the module file.

Change-Id: I965e031497920f527962ba187bccd81de6977b8f
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7336
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
2021-05-10 11:12:40 +00:00
Ziye Yang
cf41deae8b idxd: update the major so version for libspdk_idxd
Change-Id: I6ad71a65244526e99a36920c630096cc9739d94d
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7809
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
2021-05-07 15:03:44 +00:00
Ziye Yang
82e4bfd346 nvme/tcp: Change the type of recv_pdu to pointer.
This is prepared for using the hardware offloading
engine in accel framework. And some fields in nvme_tcp_pdu
needs to be DMA addressable.

Change-Id: I75325e2cd7ff25fe938bea0ac9489a5027e3e0e9
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7770
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:41:24 +00:00
Ziye Yang
f8ac678ec3 lib/nvmf: Change the pdu_in_progress in tqpair to pointer.
This is used to prepare using the accel framework to calculate
the crc32 because some fields in this structure needs to be allocated
in DMA addressable memory.

Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: Ib8def5596e60f4702709da647145c4e2b6d6848f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7767
Community-CI: Broadcom CI
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: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
2021-05-07 11:41:24 +00:00
Jim Harris
de6e5c30c9 scsi: use SPDK_TRACE_ARG_TYPE #defines
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ifc2a79891ab26643a2783d0e6482d58449a300f5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7764
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:35:38 +00:00
Jim Harris
40cf86f20a nvmf: use SPDK_TRACE_ARG_TYPE #defines
Note: a lot of the TCP and FC trace registers were
specifying '1' which means the arg type is a pointer,
but in reality it is always passing 0 for arg1.  So
this patch just changes them to SPDK_TRACE_ARG_TYPE_INT.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I18d3cedd21e516f16cb2cd0a7f8c16670b1895d7
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7763
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:35:38 +00:00
Jim Harris
a7ea7cbc57 iscsi: use SPDK_TRACE_ARG_TYPE #defines
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I033aa060feb0f6e4a9abedd8cf1335b25437fa99
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7762
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:35:38 +00:00
Jim Harris
469a504bae bdev: use SPDK_TRACE_ARG_TYPE #defines
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Iddcc9ab7b5077368a4fbda72e650b381ef42b6cb
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7760
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:35:38 +00:00
Jim Harris
a4c37c6a56 ftl: use SPDK_TRACE_ARG_TYPE #defines
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I10a0a2bf3f2bdcef0b4e4c76ee6958151c2b61e6
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7759
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:35:38 +00:00
Jim Harris
b8fb7d67be iscsi: track PDU object for WRITE_START/WRITE_DONE events
We already pass the PDU as arg1, so by changing the
trace register descriptions, we can map the PDUs to
more readable IDs when running the spdk_trace app.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Iad7106eeb0f5fe738f81da5ee174515d1cf4b6ce
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7757
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:35:38 +00:00
Jim Harris
7e52eb740a iscsi: fix ISCSI_READ_PDU spdk_trace_record
This was accidentally moved to the wrong place as
part of some earlier iSCSI refactoring.  This trace
record should be executed when we have finished
reading all of the data for any PDU, not just those
with immediate data.

Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ib1d17e5e79ff220e9e9b3dd55e247e745bd58019
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7756
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 11:35:38 +00:00
John Levon
bbfbcc1967 vfio-user: add disable-mappable-bar0 flag
It can be useful for testing or development purposes to force clients to write
to doorbells using vfio-user messages instead of directly into shared memory;
add a transport-specific option to disable the shared mapping.

Signed-off-by: John Levon <john.levon@nutanix.com>
Change-Id: I7ed062fbe211ba27c85d00b12d81a0f84a8322ed
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7554
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
2021-05-07 11:31:20 +00:00
Changpeng Liu
4f1530c9eb bdev: add bdev_unmap_should_split() function
Change-Id: I2680eef93da13d8378527349cdc5ba36859f89dc
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7517
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
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>
2021-05-07 11:30:55 +00:00
Changpeng Liu
d4d5e20d29 bdev: rename bdev_io_split to bdev_rw_split
This will help us to add unmap split function, also
remove bdev_io_type_can_split() because we changed
to use swith(io_type) ... case now.

Change-Id: I449d6a9f5bf2d0b43dd124bbfc9e1ca2afddc15a
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7516
Community-CI: Broadcom CI
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>
Reviewed-by: <dongx.yi@intel.com>
2021-05-07 11:30:55 +00:00
Changpeng Liu
518d28dba9 scsi/bdev: free bdev_io when there are split requests
Only the last bdev_io can be freed without this fix.

Change-Id: I0d05b5d89e38ef60872ebc0f23aaed0c622593c4
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7571
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: <dongx.yi@intel.com>
2021-05-07 11:30:55 +00:00
Ziye Yang
20698a4a8d idxd: Split the idxd library into common and user space part.
Purpose: This patch is used to prepare to add the kernel
idxd support later.

Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Change-Id: If89665f95d622c7342ab75050664158ec6fc615a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7330
Community-CI: Broadcom CI
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: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
2021-05-07 08:56:20 +00:00