Currently SPDK NVMeoF will allocate data buffer from mempool, however,
some data buffers may not block size aligned, this is okay for NVMe
backend but will fail for AIO backend. 4KiB aligned data buffers
should be okay for most AIO backends.
Change-Id: If7b9c2962d716addbd755e80e59620377f2ba488
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/389245
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
There are two ways to set stockets to nonblocking type:
- ioctl with FIONBIO
- fcntl with O_NONBLOCK
Those two should be equivalent for sockets used in SPDK.
During testing it was shown that VPP interprets only
the second type, so this patch changes all occurences of it.
When here, more descriptive error logs were set in case of
failure.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Ifa5b30e3a4fa04fe23f41fa2ae9dab4b01dd7d3c
Reviewed-on: https://review.gerrithub.io/388816
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
We probably do not allocate the mgmt_channel when we destroy the
rdma queue pair, add the judgment before put the io channnel.
Change-Id: I528c9686907c04c5025ff9d3872f24f2996e9d33
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
Reviewed-on: https://review.gerrithub.io/388159
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
According to the current code, rdma qpair is always created
by the thread on acceptor_core, thus we need to
change the related I/O channel during polling if the core
configured for subsystem is not same with the acceptor core.\
With this patch, we can run NVMe-oF tgt with multiple cores,
and each subsystem can configure different core to handle.
Change-Id: I6163a871f65115e545a4f3fd9cc46b3bafb13249
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/383683
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ibdfa8be770d7bfcb2baaf29fa5b32dea064ffbd0
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/383383
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I65e84971f2d55f27b0c0c1a1b226fc4da4b3cf89
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/382763
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Our previous implementation makes the buffer allocation
in an unfair way and this patch can solve this issue.
With this patch, we can use limited buffer to support
high I/O depth.
Change-Id: I0e7a073c0b4539090218aa461d50620287bb4b63
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/382528
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
For the normal case, we can avoid re-adding it into
pending data buf list if there is no buffer.
Change-Id: I400e3c6c60bb1d1df13782faf2dd32ca2bd897a3
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/382402
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This is not a public API, so simplify the number
of internal header files where important types
are defined.
Change-Id: I115d0497d37e3cfe399c3a5b2546d20aa4fe24b4
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/376249
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
We are still creating one poll group per controller,
so this isn't particularly useful just yet.
Change-Id: I65c54385bdba587d4b1098629727877970d39277
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/376241
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The requests are now sent using spdk_thread_send_msg
internally, so the user no longer needs to supply
these callbacks.
Change-Id: I84b0f5f0f1f6fa6eaf9a717934925d3ad802fcfd
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/376240
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: John Kariuki <John.K.Kariuki@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Instead of polling each individual qpair, polling is now
done by poll group. This allows transports to use more
efficient polling schemes in the future.
The RDMA transport as of this patch still just loops
over each qpair in the group and polls it individually,
so this patch results in no performance change yet.
Change-Id: I0f63f0dbbc5fd43c1e0d9729b10b37c2cb0d9881
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/376239
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Currently they're entirely contained within a single
spdk_nvmf_ctrlr, which won't span devices, but this
sets the stage for a more flexible library.
Change-Id: I653f3d6fe4187f4eaf18cda0a6960040ba6952d7
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/376238
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Everywhere this is used, we can just use
spdk_nvme_transport_id instead.
Change-Id: I407c812cd6a1ca3afcb3d1fe87c7e22183b09d7a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/375461
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This matches the name to the behavior and prepares for addition of a new
log macro for "info" log level.
Change-Id: I94ccd49face4309d3368e399528776ab140748c4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/375833
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The list of options is going to continue to grow, so avoid
constant API churn by adding an options struct.
Change-Id: Ie9e7248281726d4aee42b3519fcf7535ba01ee34
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/374872
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Formalize a state machine around request processing.
The state is advanced by calling
spdk_nvmf_rdma_request_process().
This clarifies the implementation considerably and
cleans up a few corner cases. Unfortunately, the diff
is also enormous and there does not appear to be a
way to reduce it.
Change-Id: I5741da24bcffc1aef367ebfe3dd1f589c5746901
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/374540
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
This function also looks suspiciously generic and can
probably be moved to general nvmf code in the future.
Change-Id: I7849d2d7952b245a065393ce4be5d5e05c8bd8c2
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/374539
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This can probably be moved to the generic nvmf code
in the future, since nothing about it is RDMA specific.
Change-Id: I99103e174d99a619d36e519820a8a8a9b56473ac
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/374538
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Previously, the shared buffer pools were allocated on the
nvmf controllers. When a new connection was established,
the CONNECT command needs a 4k buffer, but we didn't know
which nvmf controller it belonged to until after the
CONNECT command completed. So there was a special case
for CONNECT that used in capsule data buffers instead.
Now, the buffer pool is global and always available. We
can just use that always, with no more special cases.
This has the additional nice side effect of allowing
users to run the target with no in capsule data buffers
allocated at all.
Change-Id: I974289f646947651c58d65cf898571d80e9dee9b
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/374360
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
This is old debug code that isn't needed anymore.
Change-Id: I2c9ab9b3a63178169c48342bfacb39c59f261087
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/374359
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
The memory maps are device specific, but the buffer pool
can be shared across the entire transport. The size
of this global buffer pool isn't currently configurable,
but needs to be.
Change-Id: Ia32a11c7de5b63250bc22e80ccfe33d780e9bd5e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/374373
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I82afad8a8607d4cd12e2e13c763fb5ecb041bb4a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/373663
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
replaces all references to strerror in the spdk lib directory with
references to the thread safe strerror_r
Change-Id: I80d946cce3299007ee10500b93f7e1c8e503ee41
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/374012
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Change-Id: I6a733ab00b61866d68d584a330020b9cd29e25b8
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/373840
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
It wasn't used anywhere.
Change-Id: I1ea580af374c730d8e6336a6a3db6c05116e6dba
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/373839
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Port is a more commonly used term.
Change-Id: Ia85a49b8e36ab73d777a905081f1584811b8cdec
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/373838
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
There is one of these per ibv_context.
Change-Id: I660fed954c5a77fed9a6bda2321583ef05b30439
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/373837
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The transport API now allows for multiple transport
objects, so allocate them on demand instead of using
a single global.
Change-Id: I5dd35f287fe7312e6185c75ae75e2488ec8cc78e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/371990
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
For now, this is a name change of controller to poll_group
in the transport layer. Over time, the poll_group will
become a more general concept than a controller, allowing
for qpairs to be spread across cores.
Change-Id: Ia92a2934541ad336f462f73175d53aaaf021f67b
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/371775
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I10857de922d5a17131910aca92c73995ea6ab8f6
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/373828
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
There are now three simple functions on the transport:
listen(transport, trid)
stop_listen(transport, trid)
accept(transport)
This makes the code quite a bit simpler.
Change-Id: I550343a084b5c095240703952c8c07ae535b5c16
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/371774
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Create one transport per nvmf target. Today, there is just
one global nvmf target, but this paves the way for multiple.
Change-Id: Iaa1f8c5e7b3c1e87621ef2a636c68c2dd8fd929e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/371748
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Match the terminology used in the NVMe-oF specification,
which is queue pair. For the RDMA transport, this maps to
an RDMA queue pair, but may map to other things for other
transports. It still is logically a "connection" from
a networking sense.
Change-Id: Ic43a5398e63ac85c93a8e0417e4b0d2905bf2dfc
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/371747
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is just a rename - the functionality hasn't changed.
Use the same terminology as the specification (which is controller)
so those familiar with the specification can more easily
approach the code base.
This is still conceptually equivalent to a "session" in the
networking sense.
Change-Id: I388b56df62d19560224c4adc2a03c71eae6fed0d
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/371746
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
The NVMe-oF target was written before we defined
spdk_nvme_transport_id. Now that we have it, go back
and replace all of the locations where we individually
tracked traddr, trsvcid, trtype, etc. and use a trid.
Change-Id: I84334a12c7581f414c1e84680f122fe885a3b9dd
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/370744
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Change-Id: Ie05f58e677107072fea6cc7702bab47a077cb595
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/370743
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
This allows the user to optionally specify the address family for
construct_nvmf_subsystem (default is IPv4).
Note that the RDMA transport still only supports IPv4 because of the way
it binds to the listen address; this will be fixed in a separate patch.
Change-Id: I534ed75f6f81e53559d1bebcd2f34f1a2b210a97
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/367429
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
- rename spdk_malloc_socket to spdk_dma_malloc_socket
- rename spdk_malloc to spdk_dma_malloc
- rename spdk_zmalloc to spdk_dma_zmalloc
- rename spdk_realloc to spdk_dma_realloc
- rename spdk_free to spdk_dma_free
Change-Id: I52a11b7a4243281f9c56f503e826fd7c4a1fd883
Signed-off-by: John Meneghini <johnm@netapp.com>
Reviewed-on: https://review.gerrithub.io/362604
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Replace with it with check the returned req
via spdk_unlikely macro
Change-Id: I1202b3955af9a68496d8ced7cf66c20cf26f7fff
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
When data needs to be transferred from the controller
to the host, do a single ibv_post_send containing
both the data and the completion.
Change-Id: I072c545b31593e0e324c97ed700b42c6a4c358e1
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This call had been reduced to a simple wrapper
around the ibv call. Delete it.
Change-Id: I42926d123db262617119a9cff77bc0d0eb1e8f31
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
These functions were only called from one place and
their functionality has been reduced to a wrapper
around the underlying ibv call. Remove them.
Change-Id: I65182012dbe6393b9d57f4191fd327bcd025a6c8
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This keeps all SGL handling in the prep_data function.
Change-Id: I9bfeed3748c1b329288350b85aa87bd604cfce4e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Now that all of the SGL mappings are static,
this function just called ibv_post_recv. Delete
the function and call ibv_post_recv directly.
Change-Id: I45216170a157709249b08c4cb0ebdb1adb906049
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
For an NVMe read, send the completion immediately
following the RDMA WRITE, without waiting for
the acknowledgement. RDMA is strictly ordered,
so the WRITE will arrive before the completion.
Change-Id: I7e4e01d7a02c2130b655ef90f5fdaec992d9361a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Except for a CONNECT capsule, always use the central data
pool for RDMA READ/WRITE operations. The in-capsule
data buffer is associated with the receive operation
while the pool data buffers are associated with the
completion, and using the in-capsule data buffer
causes a lifetime mismatch.
Change-Id: Ieb45e521d78daa7c706078a3dd5c5a146f8dc1d6
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
After commit b654e9b, this is no longer required.
Change-Id: I0cf1a7059d7fba0303aca5ad5a15afe3890b4172
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
