Put clear behavior of setting flags to funtion vbdev_ocf_mngt_exit.
No need to add an extra funtion any more, so remove it.
Signed-off-by: yidong0635 <dongx.yi@intel.com>
Change-Id: Icb5b569ada597b4896523c95ba19abae3f064f4a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/2384
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Put these exit behaviors into one function.
Make this more readable.
Signed-off-by: yidong0635 <dongx.yi@intel.com>
Change-Id: Ifd7a74910847dffd3d8c6c6db61e6e1fc7a8d779
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/2283
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
The spdk_opal_supported() is redundant with spdk_opal_dev_construct(),
because we only return the spdk_opal_dev structure when the drive
can support OPAL feature.
Change-Id: I2a8b70aa92828cf60d168dcf2985759e0eb9a6e3
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/2217
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
IO device "bdev_nvme_poll_groups" is not unregistered in the finish
module, so we need to unregister this io device, since the "nvme_if"
is staic variable in bdev_nvme.c, it's can't be accessed in common.c
so we use "g_nvme_bdev_ctrlrs" instead.
Change-Id: Id4675f5acc8f386609903497da5ca84bf8af3a15
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/2035
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Seth Howell <seth.howell@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Purpose: To set the priority of the NVMe-oF connection especially
for TCP connection.
For example, the previous example can be:
trtype:TCP adrfam:IPv4 traddr:10.67.110.181 trsvcid:4420
With the change, it could be:
trtype:TCP adrfam:IPv4 traddr:10.67.110.181 trsvcid:4420 priority:2
The priority is optional. We try to change
spdk_nvme_transport_id but not in spdk_nvme_ctrlr_opts since
the opts in spdk_nvme_ctrlr_opts will reflect in every nvme ctrlr,
this is short of flexibility.
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Signed-off-by: Sudheer Mogilappagari <sudheer.mogilappagari@intel.com>
Change-Id: I1ba364c714a95f2dbeab2b3fcc832b0222b48a15
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1875
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI
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>
Last and only usage of struct lvol_task was removed in
previous patch.
Since it is no longer used, remove the structure itself.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I3612106de2ab3a9197245de078c73c53beeff443
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1971
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>
There is no point in setting this value on submission
and changing it on callback.
Since it is based on blob bserrno.
lvol_task is removed in next patch.
This change was motivated by lvol_op_comp() accessing
the task->status pointer on hot path for I/O.
There is no need to do so, so just pass status.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I17461fcf355f771f4de4e6d5ec1900d5042bd3d5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1804
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>
This is always going to be true because of the
context we are in when calling this function, but
add an assert just to be safe.
Signed-off-by: Seth Howell <seth.howell@intel.com>
Change-Id: Ifbd075865a191529eb9e656c360f182dba8a85bc
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1985
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: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This patch adjusts several return codes to provide
more than just -1.
Along with fix to json rpc error print,
where negative error code was passed to spdk_strerror().
Resulting in unkown error being reported.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I254f6d716d0ce587f88cc658163ba049378f3b2f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1915
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This was changed to better facilitate thread safety.
In next patch a lock will be held when going over the
cuse devices list.
Now user is expected to pass a buffer of a sufficient size
that will be filled with ctrlr or ns cuse device name.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I3202ef285e427111e3595389619463fda58dbef6
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1978
Community-CI: Mellanox Build Bot
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>
Unregistering nvme_cuse when the device did not exist
resulted in SEGFAULT within nvme_io_msg_ctrlr_unregister().
To prevent that, when no nvme_cuse is registered for the
ctrlr do not unregister nvme_io_msg_producer.
RPC and spdk_nvme_cuse_unregister() now return an error.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Id77cebe23ff91023a24cfe091f5f62a76a9175fd
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1921
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI
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>
All of the error cases for spdk_nvme_transport_id_parse_trtype
were covered earlier, so just make this an assert.
fix kw warning 12653.
Signed-off-by: Seth Howell <seth.howell@intel.com>
Change-Id: If0eea934764b35d097887b98b0441ae5d7dd3e4f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1953
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>
New OCF version is ready, so update our submodule.
There were a lot of API changes so
we need to adjust our bdev/ocf and lib/ocf_env code.
Description of code changes:
- env: added small functions like ENV_BUILD_BUG_ON, env_vzalloc_flags
- env: added "destroy" method for all locks
- env: added "execution context" interface
- cache/core "id" field made private
- signature of "ocf_io_configure" changed
- "io->volume" field made private
List of changes to OCF is available at
https://github.com/Open-CAS/ocf/releases
But no new features were implemented for our
OCF module since 19.6
Change-Id: I1ac2b6db0aad2a9840e4fa73d65608c5638fcb23
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1801
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI
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>
I reversed the arguments to abidiff when I first submitted
this test which was causing additions to be reported as removals
and vice versa.
Signed-off-by: Seth Howell <seth.howell@intel.com>
Change-Id: Iac50243f7cda7c1f9c2302af13a34eca844e0689
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1705
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.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: Paul Luse <paul.e.luse@intel.com>
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
We will be create fine name for each poller but it will need large
effort. Replacing spdk_poller_register by the macro SPDK_POLLER_REGISTER
will provide better name than function address with minimum effort.
Following patches may improve function name for clarification.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: If862a274c5879065c3f7cb04dcb5ca7844523e68
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1781
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: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Maciej Szwed <maciej.szwed@intel.com>
Community-CI: Broadcom CI
Cache the return value of spdk_io_channel_get_thread() to orig_thread
and use it in the function. This will clarify the logic and improve
readability a little by reducing line length. Besides, in
vbdev_compress_submit_request(), spdk_io_channel_get_thread() can be
replaced by spdk_get_thread() and do the replacement in this patch.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Id0fadb5dae9b78ecade99c5deb650849bf6257ab
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1453
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Previously only bdev_close() had been ensured to execute on the
correct thread.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ie3019bd1206060a86ec3c598795255c4c4357b0f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1696
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Community-CI: Broadcom CI
Not only bdev_close() but also TAILQ_REMOVE and bdev_module_release_bdev()
are better to be executed on the correct thread for
vbdev_compress_destruct_cb().
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I0810ecc4149747315e6ec290437ba40fb4b71686
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1694
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
vbdev_reduce_init_cb() had ensured only closing bdev was done on
the correct thread. I/O channel should have been released on the
same thread.
Fix this potential issue. reduce_errno and vol cannot be passed
through message because message can have a single parameter.
So set vol before sending message and use vol as reduce_errno.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I923a9e3148db0f5413a6903d62f869eec122dcca
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1693
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Factor out setup source and destination mbufs operation from
_comppress_operation() into a helper function _setup_compress_mbufs().
Setup source mbufs and setup destination mbufs are identical except
for total_length, and so this refactorinng is possible.
This refactoring will improve readability and clarify the logic
a little.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ia0a230c2d5e74aec917580aae969b49266178191
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1683
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
This request was getting leaked previously.
Change-Id: I6e7dc24c92d7bae7a8df54025d61eb391a569df9
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1591
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>
If thread A is the first to grab a channel, the poller will be on thread
A. But then if the shutdown happens such that thread B is the final
channel to release it's channel, it will attempt to free the lun poller
from thread B and crash.
Change-Id: I354f9f28d55e5e82cb6c737c734acadbd80e283d
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1590
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>
Save cache device name in metadata section to
check if names agree during auto creation on examine.
This change prevents accidental creation of ocf devices.
Consider example in which vbdev_split is used as cache:
1) bdev_malloc_create -b Malloc0 ...
2) bdev_malloc_create -b Malloc1 ...
3) bdev_nvme_attach_controller -b nvme ...
4) bdev_split_create nvme 2
5) bdev_ocf_create ocf0 wt nvmen1p0 Malloc0
6) bdev_ocf_create ocf1 wt nvmen1p1 Malloc1
7) *shutdown*
8) *startup*
9) bdev_nvme_attach_controller -b nvme
Now ocf0 bdev gets created with cache = nvmen1 (not nvmen1p0)!
This is because nvmen1 and nvmen1p0 have the same metadata.
The result is rather unexpected,
so we would like to verify that cache device
is the one used during creation.
Name is not an ideal way to verify that,
but it should work for most cases.
This change is related to issue #1320
Change-Id: I19ce97163090c5c5b3324eaed90b3134011bb3fb
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1544
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This change is aimed at fixing github issue #1312 where we are
apparently calling the nvme_bdev_ctrlr_destruct function more than
once.
In the previous implementation, if a controller was resetting for more than
one iteration of the poller, a second iteration of the poller would get
registered.
Signed-off-by: Seth Howell <seth.howell@intel.com>
Change-Id: Ica01ecb21d76ea4f60624efd780bf4eea957c277
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1611
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>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reorder vbdev_redoce_load_cb() so that normal case comes first,
orphaned case comes second, and other error cases come at last.
This will simplify the code and improve readability a little.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: Ieb79364f0b3996600e8ff7ec340101877527cac2
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1477
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Move queue insertion to the end to avoid removal in error cases.
spdk_bdev_module_examine_done() is duplicated with the caller in
error cases and so remove it.
I/O channel is not got in this function and so remove
spdk_put_io_channel() in error cases.
move up spdk_bdev_close() to call at error_bdev_register or
error_claim label.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I34a1459dcb0d3e0a2018f5b50db5be3ac44b6049
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1469
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: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
In comp_bdev_ch_create_cb(), list head comp_bdev->pending_comp_ios and
comp_bdev->queued_comp_ops should be initialized only when
comp_bdev->ch_count is 0. Besides, extra brackets for if condition
can be removed.
In _comp_bdev_ch_destroy_cb(), comp_bdev->ch_count is not necessary
to check if it is 0 again.
Fix these.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I71786423a77819c41419c51a4cfcad71b85a0b39
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1444
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
When reduce volume is unloaded at compress bdev removal, the callback
to unload did not get base bdev's I/O channel on the same thread
that opened base bdev.
Hence spdk_put_io_channel() hit assert later.
This patch fixes the bug by sending message to get I/O channel.
Fixes issue #1307.
Besides, add assert to _delete_vol_unload_cb() to detect the unexpected
complex cases. Fix will be added later.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: If4983c2e06d7b0b7618f38fb80f3aa73effe4b83
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1426
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: Paul Luse <paul.e.luse@intel.com>
Introduce context structure to ensure the callback function to
bdev_compress_delete() runs on the original thread.
The time to free comp_bdev and the time to call the callback function
to bdev_compress_delete() are different. Hence it is necessary to
allocate context separately.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I534f5673b58d2a0321ee752534c3ea10fb37b11d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1472
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Commit 8c1d107 added a new erase API and renamed the original
API with "secure_" prefix according to the specification, as the
secure one will not cryptographically erase user data, due to
the limitation from the drive, only the secure API can be
supported, so changing the caller to use the original again.
Change-Id: I25cb054e728065d9ae9c6c30aa0ab374d60cf6f6
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1519
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>
Actually we can just use one API to finish the requirement.
Change-Id: Ia0d3d589755e8c92f636d3d090ec642299511401
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1280
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This will eliminate one calloc() and add one small code cleanup.
Change-Id: I3d67f2b073a46e04304ae5342974afecf532794b
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1233
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The revert asynchronous API doesn't run as the *real* asynchronous
way, because the drive can only support synchronous module and only
1 session is supported. The reason why we added this API is that
RPC call has the default timeout value here, while the revert may
take over several minutes, the API itself doesn't short the revert
action, so just remove it and use the synchronous API instead.
The revert action will erase all the users data and bring the drive
back to the factory state, it should run in the synchronous mode,
so just remove the asynchronous API and we can increase the timeout
value when using RPC to call this API.
Change-Id: I08a082edea6385e378399423bbb229d05f8bc262
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1232
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Sending large amounts of data from host to the controller with 1s
admin polling time, take a lot of time (e.g. 1M firmware file in 4k
chunks takes ~17min).
Reducing this time to 10ms whole operation takes about 3s.
Change-Id: I2dabe9f60acab57e348c34bfabc3cc7479dedec9
Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1393
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This will allow us to keep track of compatibility issues on a
per-library basis.
Change-Id: Ib0c796adb1efe1570212a503ed660bef6f142b6e
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1067
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>
Default was 8 which meant max of 8 bdevs. Bump it up to 64.
Fixes issue #1232
Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: I966e90de5c27910df0e4da0d1062d9d1665f8de6
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1063
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>
The adminq poller could get a failure if the ctrlr has
already been hot removed, which starts a reset.
But while the for_each_channel is running for the reset,
the hotplug poller could run and start the destruct
process. If the ctrlr is deleted before the for_each_channel
completes, we will try to call spdk_nvme_ctrlr_reset() on
a deleted controller.
While here, also add a check to skip the reset if the
controller is already in the process of being removed.
Fixes#1273.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I20286814d904b8d5a9c5209bbb53663683a4e6b0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1253
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>
This isn't in the performance path, so using the mutex
here makes it a bit more consistent with other ctrlr
members such as 'destruct'.
This prepares for a future patch which will defer
ctrlr destruction on removal if a reset is in progress.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ica019cd90dc3b46ef6a13dd311054dbdc95855aa
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1252
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>
Make sure a namespace is depopulated only once all outstanding commands
generated by the media manegement poller to that namespace are completed.
It fixes the use-after-free errors reported by asan during shutdown.
Fixes#1251
Change-Id: Iab99b594756cee2d41235c70194bcaa5f5e1fb0b
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1199
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: Aleksey Marchuk <alexeymar@mellanox.com>
Now that a namespace can be depulated asynchronously now, the NVMe bdev
module should also be finalized asynchronously, after all namespaces and
controllers are deallocated.
Change-Id: Ic082fec8e31e9bd5ee1c698cd8dfca9f248776d3
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1198
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>
