This allows the target to poll for internal completions
at higher priority.
Change-Id: I895c33a594a7d7c0545aa3a8405a296be3c106fb
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This ensures that the data buffers are not in use
when we go to send the completion.
Change-Id: I30467b3e3964001150f81b21e5b695dcd0974b0c
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This is useful for holding session-wide buffer pools.
Change-Id: I7024da24b210a2205bf1e159d5935e0093b81120
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
For small SGLs, even if they are keyed and not inline, use the
buffer we allocated for inline data.
Change-Id: I5051c43aabacb20a4247b2feaf2af801dba5f5a9
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Read/Write depth is much lower than Send/Recv depth.
Calculate them separately to prepare for supporting
a larger number of receives than read/writes.
Currently, the target still only exposes a queue depth
equal to the read/write depth.
Change-Id: I08a7434d4ace8d696ae7e1eee241047004de7cc5
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
These don't actually work quite yet, but pipe the
configuration file data through to where it will
be needed.
Change-Id: I95512d718d45b936fa85c03c0b80689ce3c866bc
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
For each connection, allocate a single buffer each
of requests, inline data buffers, commands, and
completions.
Change-Id: Ie235a3c0c37a3242831311fa595c8135813ae49e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This can be used to release requests that don't
require a completion to be sent.
Change-Id: I8fb932ea8569bf3c45342d9fa4e270af5510c60c
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
PORT IDs indicate hardware failure domains according
to the NVMf specification, which means they should
indicate which transport addresses are on the same
NIC. Unfortunately, that doesn't really make sense for
IP-based fabrics because IP addresses can move. The
safest way to present this is to show all IP addresses
as part of different subsystem ports.
Change-Id: I056a50c69be70b4fbf1f896e684ce65bd792241e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The NVMe over Fabrics 1.0 spec corresponds to the NVMe base spec version
1.2.1, so we should pretend to be at least that new.
Change-Id: I36fc44c780de01d6c666e87b803cd47dba0e74c5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These belong in nvme_spec.h anyway and are not used.
Change-Id: I889dfebee523dc5ae503fd0370bb800f1d17fb5d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is a leftover from a previous controller numbering scheme that is
no longer used.
Change-Id: I3058802f0324b0e38708111634ee993c6e884087
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Move the ctrlr and io_qpair out of spdk_nvmf_subsystem, package them
as a new data structure. Union the direct and virtual mode namespaces.
Change-Id: I839aee3372c6c57aa03a0be76f8aaeb5045ecdaf
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
CAP.CQR indicates whether contiguous queues are required; this is
meaningless in NVMe over Fabrics, since queue creation is handled
implicitly for each connection, but the spec requires it to be set to 1.
Change-Id: I6b05954eefa6928beecd7a640bbbdbd835c6b69a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Use the size of the applicable structs directly.
Change-Id: I4a65de548d409c9962b11a75d3fde2bfe434a3ec
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
nvmf_create_subsystem() already copies the name, so the strdup() in the
caller is unnecessary.
Change-Id: I225f0f077fee30051b197a4b1d7276b113ec6b01
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It isn't actually necessary to drain the cq before
destroying it.
Change-Id: I6f77ae578176a14b5de935274a14cfd165229ec5
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This logically belongs inside the session handling code, not
in the transport-specific layer.
Change-Id: I93b2271f38dbfc742162c98c40acb153c7e9022a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Track and print out the currently outstanding I/O in debug
mode with rdma tracing enabled.
Change-Id: I0a1f0cd6e22dbf21e18ca0ec7d0c2c6d194509e3
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Instead of reimplementing handling for checking the
completion queue, nvmf_rdma_accept can now call
the general purpose poller.
Change-Id: Id2c899d1e500a8cb8491e51cc101a1bf0e167764
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
AER breaks our current model of requests/completion pairs.
Temporarily handle it by immediately re-posting the
capsule while we work on a real solution.
Change-Id: Ie7a4d88030b6fff5a11c4697eec0f024f9737f27
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Inline this code into the places that called it. These two
spots will be combined into a single path in a later patch.
Change-Id: Ice2f009ad56b783dc28ebbf1abbb877ce6000293
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This is an RDMA-specific operation, so hide it inside
the transport-specific layer.
Change-Id: Iaa097e8dde78d820547b3a39e9717c992581340b
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
These can be done at the same time now that the queue depth
is known ahead of time.
Change-Id: I7ecef30ebb4311e0a1c88f37461d34534f8600bf
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Calculate queue depth into a local variable without
touching the rdma_conn.
Change-Id: Ie804ed39ddecbf59015a4e4f7aa127f1381d9080
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Make sure the trace history that is exported via shared memory is always
the same size, regardless of DPDK configuration.
Also removes the necessity of including DPDK headers from spdk/trace.h
(so we have to fix up other files to include what they use).
Change-Id: I32f88921fd95c64a9d1f4ba768ae75e2ca5d91da
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It is not currently configurable, but this will allow us to make the
discovery subsystem have config options (e.g. which lcore to run on).
Change-Id: I788a64ba4462b023453191e509ce8de59fd90ae4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is a much simpler approach and is only slightly
less efficient.
Change-Id: I909de376d576a74156c1be447e90e7dbc240f025
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Drop the redundant controller ready check.
nvmf_process_io_cmd() was checking CSTS.RDY, but this is not necessary,
since its only caller, spdk_nvmf_request_exec(), is already checking
CC.EN, which always matches RDY in our virtual controller
implementation.
The initialization of status is a dead store -
nvmf_complete_cmd() always writes the full response, and the only other
branch is the return immediately below the call, which also sets status.
Change-Id: I1ec2b8a225a91c4b2997d8ab4f45d050cc216de3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
No reason to use DPDK in this file just for an equivalent to assert().
Change-Id: Ic6932a16d0a36cd1a3cb25c8cc5e295c59f3e2db
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Temporarily set the in-capsule data size to the maximum data transfer
length. This should actually be updated by the transport layer, but for
now, the only transport (RDMA) supports the full bounce buffer size.
Also drop the check that prevents admin connections from using
in-capsule data; the host may send in-capsule data for the Connect on an
I/O queue, and we don't know the type of connection until after Connect
is processed.
Fixes: 828dca7 ("nvmf: Move some stray session init code to the right place")
Change-Id: I369ee5497247d7e875ad0b6f0aaf6c47c1d3887c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Make sure no response fields are left over from the previous command in
the spdk_nvmf_request.
Change-Id: I42937e991d9dd6550fd4bc9b6d0dd66b44c6b83e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The kernel driver unloading/loading code is Linux specific; replace it
with stubs on FreeBSD for now.
Change-Id: Ic67c1d89b2fb9a65e9ce5b88d27b6cd6af5554a7
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
spdk_nvmf_request_complete() always sets CID to the value in the
command, so there is no need to set it in the command execution
functions.
Change-Id: Ibbe745b862e27fff7c55e553758ef093e3ef7f6d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Use the passthrough command for all Identify commands except Identify
Controller.
Also only check the CNS field of CDW10 and use the new enumerated names
instead of magic numbers.
Change-Id: Ia94f820ac85a2d6b2d0ae02659e73c53f1b1a4cd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Drop the special-case preprocessor definition for PCI access library now
that config.h is available with an equivalent SPDK_CONFIG_PCIACCESS
define.
Change-Id: I4891d0f2fd7d3eea51b767df9e594555b36265ea
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
If we connected a subsystem twice from the initiator, the second
connection will be rejected by the NVMf target, however, the previous
connection will also be impacted because we destroy the connection id
before ack the disconnect event.
Change-Id: Ib597cc68a7823524460693053898f4d6e5499eb4
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
There is no need to handle Read and Write commands separately; the
generic raw I/O command case can handle them just as well.
Change-Id: I8475eed0a20bd809c447ed2ccac0b99f6c2a9b4d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Replace use of the newly-deprecated rte_mempool_count() with the new
name, rte_mempool_avail_count().
Also add a compatibility wrapper so that builds against older DPDK
versions still work.
Change-Id: If3c44bdef4bbcf7a456a1dfa272348ccc6f35261
Reported-by: Jay Sternberg <jay.e.sternberg@intel.com>
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The host is not allowed to send normal admin or I/O commands until the
controller is enabled (via the Fabric Property Set command).
Change-Id: Ib62be3a3792fc0b36bace28b4c9afdf78dad3bcd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Only allow Connect on a new connection (one that has no associated
session yet), and only allow Propert Set/Get on admin queues.
Change-Id: Iae22379ee47b095333372e6d151a7a1509acf654
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The NVMe spec requires that the I/O queue entry size values in CC are
set before any I/O queues may be created.
Change-Id: I4f0c9a9c20411223d281993745c85a8431197961
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Track each individual bit in the Set Property handler for CC, and fail
the request if any unhandled bits are modified.
Also add handlers for IOSQES and IOCQES (I/O submission and completion
queue entry size).
Change-Id: I374dc3c15197e029ba07fd9ee1cff0e38a0a884d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It is not implemented yet, but add a message to remind us to write it
later.
Change-Id: Ic1c35a0d35f728bc63b38c334d9c622493bee967
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Property Set of CC.SHN is not supposed to terminate the session - remove
the commented-out code that was attempting to do this.
Change-Id: I1db230df9be549764287a8fd45ccdebea1d22a8b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Set CSTS.SHST = 10b to indicate that shutdown is complete, and
CSTS.RDY = 0 to match the state of CC.EN.
Change-Id: Ia651c34427526a38f22cba3910df2cf7d4bedd92
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Explicitly include spdk.common.mk at the top of all lib Makefiles so
that CONFIG options and other predefined variables are set.
Change-Id: I1e560c294fe8242602e45191a280f4295533ae44
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
There is no need to allocate ibv_sge structures within the RDMA request;
we can just fill them out on the stack right before submitting each
request.
Change-Id: I438ff0be2f6d07ffa933255c92c4ec964aa1b235
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Just return success or failure - the actual count was not used.
Change-Id: I26e7c4c6319af444d221d9b0f313fb7071733619
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
All of the WC events that we handle map back to a request, so look it up
before checking the opcode.
Change-Id: I1b70a773374f64387df0a21a4f7fd64b26534b14
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Make sure all tracelogs in rdma.c use SPDK_TRACE_RDMA.
Change-Id: Idc3d3b6654215b5ab3ee84a106e46ffd3019cc7a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These NVMf spec structure definitions are the same as the equivalent
NVMe structs.
Change-Id: I21c45973b7843e3767c48f97ec42e7b446df296f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
There was only one function and a structure declaration
left.
Change-Id: I63277b4182120e7a76a925ed0bf7378ec7c23f20
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
These can be simplified and merged into the subsystem.
Remove the concept of mappings from subsystems and replace
it with a list of hosts and ports. The host is optional -
not specifying a host means any host can connect.
Change-Id: Ib3786acb40a34b7e10935af55f4b6756d40cc906
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Make the transport responsible for filling out the fabric-specific
details in the discovery log entry.
Change-Id: I41d871c605becd557dca18f8ef7e80da66950257
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Make the core NVMf to transport interface generic and allow for multiple
transport types to be registered.
Change-Id: I0a2767a47d55999c45f788ae1318bb50af60ab4e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Change the Port configuration file entries to a new format:
[Port1]
Listen <transport> <address>:<service>
Initially, this still only supports RDMA, but the new format will allow
specifying other transports once they are added.
Change-Id: Iadfd19b91db57b571064379368dbe77204ccecbb
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Each subsystem will run on a single core, which is more than enough
to fully saturate a device and a NIC. For now, all subsystems
run on the master lcore.
Change-Id: I95340a262d70fd346fa81fe519e7d4190a369e64
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Instead of starting the connection poller immediately upon
the connect event, wait for the first connect capsule to
start the poller.
This builds toward associating all connections with the same
session with the same lcore.
Change-Id: I7f08b2dd34585d093ad36a4ebca63c5f782dcf14
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
It can be different per fabric interface within a single port.
Change-Id: If13590d7f12291499ccfd705efaf6d2b1b1d7003
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The type is already stored in the fabric_intf.
Change-Id: Icd33dd29f2fa1313329b4053892693c7ff90945d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
For now, it just contains RDMA, plus a raw byte array to allow generic
copying.
Change-Id: I02fe11f99dd8b49000de0dba991cd34c99fd7a4a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Pull out the duplicated min checks against the ibdev_attr values.
Change-Id: I774c355ba669486afde5c05c55a4ed653723db98
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Set a status code in the response capsule for each possible error case.
Also enforce CC.EN == 1 before I/O connect.
The NVMf spec requires that the controller is enabled before any I/O
queue Connect commands are allowed.
Change-Id: If56d6b4d6bedad00e9e845e77f05f715e3969f8b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Drop the debug print in conn.c that was the only user.
We still have the connect data structure when determining the connection
type, and after that point, the queue ID is not needed.
Change-Id: Ida9e170099f977ec6b84478874863c40d6f7d8a1
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The NVMf target is being refactored to split the RDMA transport-specific
code into its own file. Once this is complete, we should be able to
plug in other transports and build the NVMf target without any RDMA
dependency if desired.
To enable this, change the CONFIG option to RDMA; it still controls
whether the whole NVMf target is built for now, but once the RDMA
dependency is actually made optional, we will be able to build the
generic NVMf target code without libibverbs installed.
Change-Id: I8cd90a9aaa85dcefcc9b0f8f2e7b6af21958b2a8
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Move the configuration file parsing for subsystems
into the configuration file parsing file.
Change-Id: Ie16e73cdc65fae7f2f3c3b22f9cba7f167024fa1
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The code for parsing the configuration file still
referred to a host as an init_grp, so fix it.
Change-Id: Ifa250b09de495dd7d393ccc3557fd6d56a54e790
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This never really made sense, so replace it with a list of
subsystems.
Change-Id: Ie7a9400083c091ac7142d01c23948200f515bdf7
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This is just extra complication for no real benefit.
Change-Id: I528af98e799d0641e753390fe35ff561fa3d7d76
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This allows the custom UIO hotplug driver to be used as well as the
uio_pci_generic driver.
Change-Id: Ica3316ed716827ad305eb4a146d0864d61ff190f
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Linking this new library allows applications using
the framework to be killed via an RPC call.
This only works if the RPC subsystem is loaded.
Change-Id: Ifcf91c212add620fe410589eba5490337c635776
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Linking this library into an application makes the log
configurable via RPC. This only works if the rpc subsystem
is also loaded.
Change-Id: If1340cf2a845ef159290232c26f341150c98fb9a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Use the number of devices returned by ibv_get_device_list() instead of
stopping at 4.
While we're here, drop the unused MAX_SESSIONS_PER_DEVICE definition
too.
Change-Id: I21ca6c6c95b7f2cccc1de4d0a34b95217a522bfc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is the only file that calls it, so it can be static.
Change-Id: I47573b7b38b40ad37e758234245eedbe94ae0a12
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These were internal-only APIs; initialize just checks to see that the
pool was initialized (which is already checked internally), and shutdown
just called spdk_nvmf_shutdown_nvme(), which we can call directly.
Change-Id: I95e1b912d61a38fa9934f58df7b1512678303452
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These can be isolated in rdma.c rather than being part of the generic
transport API.
Change-Id: Idc2b969a2f7685420cda2f7c4aa12495ffc3fcbc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Just calculate the required number of requests once and store it in a
global variable.
Change-Id: Iffeb637a3ac5f69ec89989b84f03699bac483b6e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
There can be only one session per subsystem.
Change-Id: I8ba85a5ebd11dd71fda2a4bafa97a0935609379f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It is just a duplicate of the NVMe library request_mempool.
Change-Id: I2a5484e5d515b965503b2cfcd8d85ccfcb0dee05
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Clean up everything that isn't strictly necessary in rdma.h.
Change-Id: Ied9acbed5f5b64860eae39816cdcb74620009a79
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This essentially turns the current nesting (of RDMA conn inside NVMf
conn) inside out. Now the transport owns the connection structure and
allocates it when necessary.
Change-Id: Ib5ca84e2a57b16741d84943a5b858e9c3297d44b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This sets up the RDMA layer to be able to embed the NVMf conn inside the
RDMA conn.
Change-Id: I5e3714ac8503826504d78d06fb5eaafabd025bb8
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The whole cleanup process is now started by
spdk_shutdown_nvmf_subsystems(). Each subsystem will clean up its
session, if any, and each session will clean up its connections.
Change-Id: I9915d4547751ed4ffc4baa2c45c628698dd0b881
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The per-lcore connection counter was incremented and decremented, but it
is no longer actually read. The lcore allocation should happen at the
session level instead.
Change-Id: I7bdf1b521bfda4892304338d43fad3ed5123c494
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Nothing actually maps the shared memory region, so there is no need to
allocate the array of connections that way.
Change-Id: I3d5eca748f892e37fbb0ec52942f1c510e9f9dc8
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
There is only one controller per subsystem, so therefore
there can be 0 or 1 sessions. Change the list of sessions
to a pointer that can be NULL if no session exists.
Change-Id: I2c0d042d9cecacae93da3e806093faf0155ddd6e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Subsystems only have one controller, so cntlid
is always 0.
Change-Id: I690a1793ad3a696adbaefca856e559dd0177b11a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This was intended to track the number of NVMe device
queues per session, but there is only one hardware
queue per session. It was conflated with the number
of RDMA queues in several places as well.
Change-Id: I74a1c56a5d395dea8bee4778882821e904cebcf9
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Everything can be done when the session is created.
Change-Id: I7cb38c093b2b1b69460cabba465828eed0cec432
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The cntlid is inside the session, so no need for
duplicate data.
Change-Id: I5669ee6393807959506dfec36a7583af77386fc4
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Since we only allocate workers to the master lcore,
remove the logic that places I/O conns on the same
lcore as the admin conn.
The "right" logic would be to place the I/O conn
on the same lcore as the whole session, and this
patch builds toward that.
Change-Id: I8983b56de41062ec834b0a169ba0fa61326c466d
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Temporarily, only run on the master lcore. This makes
some temporary refactoring possible that is required
to move to a truly scalable threading model.
Change-Id: I13a2e03107a27f8ec18b023b15f653d374a137b5
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
A connection function was initializing some session data, so
move that code to the function that initializes the session.
Change-Id: I5f2d4349585cb97985a7bbd9fb8d6c66eeaa7d4e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
There was an extra layer of indirection complicating
things for no reason. This removes it.
Change-Id: I8d4e654eb17f8f6ec028d775329794f0745fb0f7
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The NVMf target set the maximum data transfer size(MDTS) to the default value
of 128KB now, and the initiator driver will read the value and set it to the
block layer, so each command sent from initiator will not runoff 128KB.
Change-Id: I1d4f259e887b2fc70c7f1c5406c07c58f7fc9b8d
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
If any completion indicates an error, we need to close the connection.
Change-Id: I50b30aa692ae121932f1baec32f713422ff415ed
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
NVMf does not have the concept of subsystem groups; the (former)
subsystem_grp files really contain structures and functions related to
individual subsystems.
Change-Id: I4b3a64de799fffb29f8685ea4908d754516815cd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
For a single poll of the completion queue, if the user
submits I/O from within their completion callback and their
completion callback is particularly slow to execute, the loop
could potentially continue forever. To support this, we
need to limit the number of completions we'll process
in one batch.
Change-Id: If6bae47e52b36347dbe5622ace68c866ee88a0b2
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Create a list of valid properties with get and set callbacks (set is
optional to allow read-only fields).
Remove handling for fields declared as "reserved" in the NVMe over
Fabrics 1.0 specification.
Also simplify the vcprop structure to only contain the required fields.
Change-Id: I14d3ddfd008c62b75fce8e64d193c87fb6f7b5ad
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is intended to be used for examples/nvme/identify and similar
diagnostic utilities.
Change-Id: Ib2f941e9af7a3fb7555865ef253742e30ccad2b5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Multiple NVMe controllers within a subsystem does not work correctly,
since we would need to virtualize the controller data, namespace IDs,
and so on. For now, only allow pass-through mapping of a single NVMe
controller per subsystem.
Change-Id: Ib2d3576d2856c46a086f38eb6bec56f3e7a73575
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Previously, we used cap_lo and cap_hi to represent the 32-bit halves of
the full CAP register. However, it is simpler to keep them in a single
64-bit structure, and is no less efficient on 64-bit platforms.
Also name the NSSRS field from NVMe 1.2, which was previously reserved.
Change-Id: I1d5d9b0dccbb12373b4aed3db29c883881d43223
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The bb_sgl must follow recv_sgl make the logic obscure.
Change-Id: I8d47477986efd8f2d4ed964ab9373b7f157af274
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
Make sure the reactor mask in profile take effect.
Change-Id: Ia471b2b88a711f05738cf93068c4f3a8c9a3039d
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
Admin commands technically don't allow inline data,
but there is nothing from preventing us from posting
a recv buffer that could handle inline data. It just
won't be used for incoming admin capsules.
Change-Id: I3e7e4406e01ab870654a166d52221c11fc0ac683
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
We need to bind to each port declared in the config file; there is not a
single global port number.
Change-Id: I41c315588078d131c32cb145d22314047505c95c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The access to the NVMf IOCCSZ (I/O Queue Command Capsule Supported Size)
field in the Identify Controller data was incorrect.
Change-Id: I23b0aa175de8e5d8a0220e9c35e0cb6868121cb5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The maximum in-capsule data size is determined by the I/O queue bounce
buffer size, and there is no point in limiting it beyond that, so remove
the need to configure it.
Change-Id: I64806516b847e819f57ac9f62a162f7a04805b57
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
4420 is the officially assigned IP port from IANA for NVMe over Fabrics.
Change-Id: I433a5ed0780d1ffd7ca6512617759d59fa5e8def
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The queue type and queue depth are not known until
the connect capsule is processed. Delay allocating more
than 1 recv wqe until then.
Change-Id: I0e68c24bc3d6f37043946de6c2cbcb3198cd5d1b
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Currently, the recv wqe is re-posted immediately. This
closes a small window where we could get more I/O
than we could handle.
Change-Id: I9b0b1f0cc526069033b9e04f170195c4fb130e37
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This is going to be used elsehwere in teh code, so
name it according to the public namign convention
and make it public.
Change-Id: Id5fd57e78e146f3235741a251bb30244d6530f2c
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This is going to be used elsewhere in the code, so
name it according to the public naming convention
and make it public.
Change-Id: I0dcb88e902c5e609fe6acd06ad06743203fcaa60
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Break out the code to allocate a single rdma request
to be used elsewhere.
Change-Id: I687ce5ec862831fed5300157bfb4bf980d22c782
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
When Debug is not defined, SPDK_TRACELOG will do nothing,
thus cmd_type is an unused variable, and will trigger the
compilation warnings. And this patch will solve this issue
Change-Id: I821f7601a16c98e514227aee2e18fbfa61928bea
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
The queue depth allowed for incoming commands is set
such that we can do the maximum number of RDMA reads
necessary. There is never a case where a READ will need
to be queued anymore.
Change-Id: I4f7e7f4a59f6358065b82f36a5e22744af210d07
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
There were 4 variables tracking queue depths. In reality,
only one is needed once the minimum is computed correctly.
Change-Id: I9bb890e92a33a3c7bd6e27cbd31d6bee7ca0cf3d
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
NVMe over Fabrics defines its own NVMe Qualified Name (NQN) format; it
does not use iSCSI Qualified Names.
Also change the default node base for nvmf_tgt to "nqn.2016-06.io.spdk".
Change-Id: I2b73c1426ef1d8c83cc2df499d79228ea61257cd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Fix the sizes of the UUID fields to match RFC 4122.
Change-Id: I1458a22579f455cde0a67ee3ce616e78d5c810c2
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This will allow removal notifications to be propagated to the library
user (e.g. for hotplug).
The callback is currently unused, but this at least prepares the API for
the future hotplug support.
Based on a patch by Dave Jiang <dave.jiang@intel.com>
Change-Id: I20b1c2dbf5e084e0b45a7e51205aba4514ee9a95
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Use the knowledge that both the source and destination of
nvme_copy_command() are aligned to emit the aligned variants of the
SSE2/AVX mov instructions.
Change-Id: I0a7e32a3bb10b9a1920cd85691b79fa7172eecb3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The function call of spdk_nvmf_check_pools can be
directly put in nvmf.c.
Reason: This pool is created by nvmf subsystem,
it should be recycled by this subsystem.
Change-Id: I49e49bcb56079fc25d26b1f5078a1808c2f8e189
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Drop the RDMA-specific fields from spdk_nvmf_request and get them
directly from the command SGL in the transport-specific read function.
Change-Id: Icd06a9018a8c341213fbc8d26d3d7cbf2fb32d30
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The connection will be closed in these cases anyway, so just let the
normal connection cleanup deal with the active tx_desc.
Change-Id: I96c68d5802e189bb82b180cc3c7d7c3f4135be1f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
If the transport poll routine fails, we need to close the connection.
Change-Id: Ie534b0f05e6642c31e0450865e309a784abbe744
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
If spdk_nvmf_request_exec() fails, the connection will be closed anyway,
so just leave the tx_desc in the active array; it will be cleaned up in
the normal connection cleanup path.
Change-Id: Ie4f60bd6001658403dd7e1c6a47d40be756ef6f2
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
If an invalid SGL is specified, send a response with a status code
indicating what the error was rather than silently dropping the command.
Change-Id: I12d1fd847d3bc0ea8de7698e934626c2586a7452
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Make all command processing functions return a bool to indicate
asynchronous (false) or synchronous (true) completion.
Change-Id: I7c2e4d28fa473b36ff26c902e4bb69f38b64d18d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Only use SPDK_TRACE_RDMA within the RDMA transport code.
Change-Id: Ie15fd24bb142a68f3661929267ebe396b556c351
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The error case could only be reached with tx_desc != NULL in one case,
so move the cleanup code there and drop the goto.
Change-Id: I7aace6b40dd75ef8d86fb173f9d58110e929b082
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Also split the generic nvmf_trace_command() function out of
the RDMA-specific handler and move it to request.c
Change-Id: If29b89db33c5e080c9816977ae5b18b90884e775
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Also finish up the req_state -> req conversion.
Change-Id: I131dd52dcd36a790b942e06f0207a3274cc04ffc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The overflow condition can't happen unless there is a programming error
in the nvmf_tgt library; we can only possibly receive command capsules
(sq entries from the point of view of the host) if we have posted a RDMA
Recv for the command capsule memory region.
This means that we also don't need to track sq_tail in the NVMf library.
Change-Id: I101509080c744528871e72fa46d188e2850c928a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The "immediate completion" cases in spdk_nvmf_request_exec() already
call spdk_nvmf_request_complete(), so the ret == 1 case in nvmf_recv()
is bogus.
Also fix a couple of spdk_nvmf_request_complete() calls in
nvmf_process_admin_cmd() that should be handled by its caller.
Change-Id: I41b865d5e6e7fec08087faf9c6f3da3b057a5fb2
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These are not supported (and did not actually function) in NVMe over
Fabrics. Queue creation is handled automatically when new connections
are initiated.
Change-Id: If3a10e5df2f0625537b2c453cd8c835e570fa31e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Move toward making request.c transport agnostic.
Change-Id: I25fbe74fff21a5c23138e1a6e2d40bc6a4a984ec
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Make nvmf_post_rdma_read() interface generic (don't require a tx_desc).
Change-Id: I331a93eed4bb1912a47a88bb904cf392fcc364c6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This fixes an oversight that allowed in-capsule data block SGLs to
potentially refer to more than the received in-capsule data size.
It also makes spdk_nvmf_request_prep_data() less dependent on the
RDMA-specific rx_desc/tx_desc structures.
Change-Id: I34d61aca4cf5ba033849673116d16ec90488dcd4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It is the same as spdk_nvme_cpl, aside from reserved fields.
Change-Id: I62b0718dd58c998b4d26a0d1b44ee16d37eff25d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The RDMA read and write commands can determine the desired length based
on the nvmf_request length field.
Change-Id: I97b63289556e7de3c19c5a17ecbacbbbdfc10425
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Replace the generic "msg_buf" naming with command and response.
Change-Id: I19baff43b41a5eb7db9be9d7feec33d17112e320
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The mempool functionality is never used at runtime - all bounce buffers
were immediately assigned to a rx_desc.
Change-Id: Ie2195059858e34b30b07e104739f046c13abc335
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The RDMA tx_desc and rx_desc pools were only used at startup; all
descriptors are immediately allocated and put into a queue, and the
mempool functionality was never used at runtime.
Change-Id: I2882274962550191a555c8483b8f7be2854b32ec
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is an implementation detail of the RDMA layer.
Change-Id: Ib97d6fbd593789eed0b6e746972b8882a3320995
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This code is operating on a list owned by the RDMA
connection, so move it to rdma.c
Change-Id: I8b81f9d1ffc1df489c9b698969725ed0d1db6a06
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
These are an implementation detail of RDMA, so move
them into the RDMA portion of the connection.
Change-Id: I68d146019c5d78fbf5e9968abfd7baed2a54a2ed
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Separate out the RDMA connection from the
NVMf connection. For now, the RDMA connection
is just embedded in the NVMf connection, but
eventually they will have different lifetimes.
Change-Id: I9407d94891e22090bff90b8415d88a4ac8c3e95e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This structure will be expanded in future patches.
Change-Id: Ibb04917134243560e09a2a255844739eb33fab65
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The request needs to know which connection specifically
it is associated with.
Change-Id: I492b9968b4d2e307b5af44edee0778478b32d2ba
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
They each only had 1 function left that belonged
in the session.c file.
Change-Id: I405902b02e9316d2dc02d3732d8bc085c2b84d31
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Only move nvmf_request definitions from nvmf_internal.h
for now. Subsequent patches will move more.
Change-Id: If47472542515fd050cc78d95540eb25beee59d2a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Fabric commands were skipping a step, so unify all
types of requests through the same completion path.
Change-Id: I5f38a7e1cdcdf33baf71486d5ddae9f5a6157fac
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The nvmf_request structure holds the pair of pointers
for rx_desc and tx_desc.
Change-Id: I3e735979bbdcdc0e70ad78762e289849d41158ba
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This moves some definitions from nvmf_spec.h to
nvme_spec.h based on the latest publication.
Change-Id: I51b0abd16f7d034696239894aea5089f8ac70c40
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Swap the order of checks in the failure check - if rc is not 0, addr may
be garbage.
Change-Id: I110710efd00397c777d59ac8b219ba3cc2156596
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The nvmf_request object is generic and is mapped 1:1 with rx_desc.
Change-Id: I397224a3859c3c93d6eca99f7ba7c53ce7963f57
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Instead of searching the global list of connections to find a matching
cm_id, we can just store the pointer back to the spdk_nvmf_conn in the
rdma_cm_id context field.
Change-Id: I39ea16be6a633a1136d65743747b63b600f20e63
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
All of the variables are private to conn.c, so they don't need global
visibility.
Change-Id: I7c24cfc6249a9f8164b162b4b8de0e24c452e0df
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It is always set to nvmf_process_async_completion and is only used
within the library.
Also rename nvmf_process_async_completion to spdk_nvmf_request_complete
to clarify its purpose.
Change-Id: Ie737fb60688329bfe329a8553c4a40ff2e5f8f1d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Create spdk_nvmf_request_prep_data(), which handles SGL processing and
data transfer for all command types, and spdk_nvmf_request_exec(), which
executes a command after data transfer has completed.
Change-Id: I51c2196260dd0686db8acca4d8f7c93e17618c2f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The pending type can be determined based on the command opcode.
This also moves the "issue pending RDMA reads" case out of the I/O queue
handling into the generic continuation code; this should not make any
difference for the current case, since the Fabrics Connect command is
the only other continuation case currently, and there cannot be any
pending RDMA reads in that case.
Change-Id: Idddfa496b6e5b7e6da772aa3ab1b9d1a5344771f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
nvmf_connect_continue() no longer needs the RDMA-specific tx_desc.
Change-Id: I95f6938063e9853aa7dcd419f488b91422ff9b60
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Let the calling function handle the tx_desc if nvmf_connect_continue()
fails.
Change-Id: I25a8cbc4c3be0608bcec8db2fb8c50e55fbe3e8c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Everything necessary for processing an admin command is now stored in
nvmf_request.
Change-Id: I74e75a5b7bb3b406ad167c2b31cab1af7a1f270a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Everything necessary for processing an I/O is now stored in
nvmf_request.
Change-Id: I3f390707ebe83ea66a116dcfda4d0388a6823629
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Keep a pointer to the local bounce buffer in the transport-agnostic
struct nvmf_request rather than groveling in tx_desc/rx_desc to get it.
Change-Id: Ic328d8e2b3a15759ccb149a89fb3562e928ca500
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Now that we have xfer to track data direction, the length field can be
populated correctly for all transfers (including in-capsule data).
Change-Id: I7b2228f3fac80aab983a4103ba095c7bc38e0b21
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This field is used to decide whether data needs to be transferred back
to the host after a command is completed.
Previously, this was determined using the length field, and length was
cleared to 0 after a transfer was completed. However, length will be
used in future patches after a host to controller transfer completes, so
we need some other way to tell what kind of data transfer is required.
Change-Id: I6b27cf7816908394735fc95c15bd5eb40a7c0157
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
vtophys() was checking for out-of-range addresses incorrectly: vfn_2mb
is already shifted to account for 2 MB hugepages, but it was being
compared with a mask that did not account for the shift. This would
allow out-of-bounds access to the 128tb map array for certain invalid
addresses (it had no effect on addresses within the valid userspace
range).
Change-Id: Ida7455595e586494c9025f9ba65d050abb16b1b9
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Other subsystems can depend on this one and then use
SPDK_RPC_REGISTER to register RPC functions.
Change-Id: I557f774331ce7146d299d06b3f81426e2103a11f
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Since we used u64 to mask CPU cores, the available number of CPU
is 64, for default RTE_MAX_LCORE in DPDK, the value is 128, in some
cases(e.g.: when nr_io_queues > 4) we can get the wrong lcore ID.
Change-Id: Icc334b1bf5b068a310839118be341e61071cff65
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
A transfer using less than the total bounce buffer size is a normal
occurrence and not worthy of a tracelog. Also drop the pointless
conditional.
Change-Id: Ibcdcf693fea439d5034fa51b08b3fbd8fd7df8f2
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Avoid using the RDMA-specific tx_desc when the transport-agnostic
nvmf_request will suffice.
Change-Id: Id35bbdfb353cb72e0feb4f5af19e5bd5c86d3ff4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Should set tx_desc=NULL here. If not,When
nvmf_post_rdma_recv(conn, rx_desc) fails,
we would make tx_desc deactive again, and
this is wrong.
Change-Id: Ieabc7e3864b7f124b003d052f66ab8799a1d632e
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
The variable cur could be null, so if we use do while
will cause a segment fault
Change-Id: I19ec26e88948a0c3fd957e03e717b68750f40c62
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
The conventional rule for returning errno is negative, hence there is no
need to modify caller's code to adjust this NVMe library.
Signed-off-by: Minfei Huang <mnghuan@gmail.com>
Signed-off-by: Minfei Huang <minfei.hmf@alibaba-inc.com>
If we just pass NULL to rdma_create_qp, it will do
the right thing.
Change-Id: I9621a5110ace6237a1e47c6e5defb4cac3afc4ae
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The wrappers are much simpler to use than the low
level ib verbs calls.
Change-Id: I4b09a96a60020bc27df9396d40d955733f618837
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
They are only ever called in sequence and do related
operations.
Change-Id: I825abe08deba1dafb405757bb4f2d52062a801ca
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This enables SPDK_NVMF_BUILD_ETC to be moved out of the library as well,
since only authfile was using it before
Change-Id: I10d1145881f9a0358d7effe2d2d9851899413e1b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
SPDK_NVMF_BUILD_ETC will be cleaned up in another commit; it is
currently used both in the lib and in nvmf_tgt.
Change-Id: Ibc5f15cc4341f9d52b29c84defcd332bec4a4d09
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Most of the #include statements in nvmf.h aren't part of the public API.
Change-Id: I0d43dd542a28744a91a4fd0c4c806a991d1e194e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It is not part of the NVMf library's public API.
Change-Id: I665d5713343c9185cbdadaef4fedfdc83b8232d6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
There is only a single global g_nvmf_tgt that can be passed to this
function, so remove the parameter and use the global directly.
Change-Id: Ia1a2a1e6cd3801101ddeb4de5526dd115fa7ef8f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The section is really defining a subsystem as defined
by the NVMf specification. There does not appear to be
any need for a group of subsystems.
This change only updates the configuration file. It does
not remove all references to a subsystem group from
the code.
Change-Id: I38e62735a5ac924dcafacb3c9a332a103d751d4a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The specification refers to this concept as a Host,
so use that term. This only changes the configuration
file usage. Initiator groups are still referenced in
the code and will be removed later.
Change-Id: I897f4dbdfb65d94da1e5a77434fc07a2c18bcdc2
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Eliminate the misconception of reactor, it actually not a subsystem.
Change-Id: I63ea46f0dfa34661f16526a71c47e8fba9813474
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
The index should be 0 for fabricintf.
Moreover, when there is no fabricintf found, error should
be returned
Change-Id: I3aa04566a5a318b8c921dd37c8573ed075254266
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
There is no logical split between nvmf.c and framework.c, so combine
them and drop nvmf.c.
Change-Id: I91230c01ed7f171bfed04456b0bfcf0e7ddbc263
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The mutex is initialized, but otherwise is unused.
Change-Id: Ia68adbd430fad391cc465c07dd6e937e90dd2c5c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The code that actually removed items from the list was removed in
addition to the free() call, which caused a hang on shutdown.
Change-Id: If0e843d0d0ebfa28638b12104da880e70b3e548a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Previously, there was no way to determine what namespace ID was assigned
when a namespace was created via the NVMe library interface.
Also drop the incorrect comment about calling
spdk_nvme_ctrlr_process_admin_completions(), since
spdk_nvme_ctrlr_create_ns() checks the admin queue internally.
Change-Id: If90a6e9fc773aefa220ebbf6effc2d033c9f20cc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This lets spdk_nvme_qpair fit in 128 bytes exactly.
Change-Id: I7c42582f22ece72a7f1d651468e63d4fe05babd6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Avoid potential double free cases.
This fixes a clang warning during scan-build.
Change-Id: I487d6fcd485d1f8ebb96b6f8cb54511628461f39
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
All trackers in outstanding_tr should have a non-NULL request. Add an
assert to verify this.
Fixes a clang warning during scan-build.
Change-Id: I0ac4d2bad17449f684808cbb98777627d890b65b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Document the requirement that qpair and cmd can't be NULL.
This placates clang, which previously generated a warning
during scan-build.
Change-Id: Ic2d5e808faee0028c890ce1312444fb3dc95f223
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These allow linear searches of the configuration file
sections.
Change-Id: I8d8b9594bc8a974c16d999689a6195434c1efac8
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The user can determine whether submission queues will be placed in the
controller memory buffer by checking the controller options use_cmb_sqs
flag in the attach callback.
Change-Id: I8a925ef99a48665a0e2ffaa90d9ff2b79b90b2fa
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Add a special "all" trace flag name to set or clear all registered trace
flags.
Change-Id: Ib579df7c41ce4aca72174e04734df20f2752035c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The D3700/D3600 series support Controller Memory Buffer(CMB) feature,
CMB is available for holding submission queues, for those controllers
which can support submission queues in CMB, user can set the option
whether to enable it or not.
Change-Id: I8b0dc9e28dd6f5bb01bee99a532087212c04e492
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
When recursively calling spdk_json_write_val() on another object or
array, the child call will handle printing out the whole
subobject/array, so the parent call should skip over all of its values.
Also fix the return value for the array/object case - if we get to the
end of the array or object, we should return 0 for success.
Change-Id: I1da80c88ab8759620114c1ab141baaaaf9f0023a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Resolve relative paths before using them to clean up command lines.
This should also help shorten the overall command line length that gets
embedded in the binary and used when locating the executable from a
coredump.
Change-Id: Ibff9849ede198bb04313496c8b7131485ffaf14f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This patch add support for Intel specific log pages :
marketing description page.
Change-Id: I87bccb2af286279598c9dd3c870094b384a0d2f7
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
For existing \file markers, move them to the top of the header and tweak
the wording for consistency.
Change-Id: Icce748effe4dbe97d79a8c87d31caf0ee5797058
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Function to trim leading and trailing whitespace from a string.
Originally based on code imported from istgt.
Change-Id: I87abe584130bdf4930098fadb8e57291f18eda7f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Parsing function for delimited strings with embedded quotes.
Originally based on code imported from istgt.
Change-Id: I448feb53ea232048ed8c68738e12bc3660eb4235
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Add function to convert string to lowercase in place.
Originally based on code imported from istgt.
Change-Id: Ica9fe2208e6ee09b22c9a652a33c5affe5be23cc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
For those controllers which can support end-to-end data protection
feature, add the support in the driver layer.
Change-Id: Ifac3dd89dec9860773c850416a6116113a6ce22a
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
When using DPDK PCI support via VFIO, the PCI device is reset
immediately before calling the PCI driver's init function. In some
cases, the device seems to not be ready to handle MMIO accesses right
away. Until the cause of this issue is fully understood, add a 500 ms
sleep as a workaround.
Change-Id: Ic893080a6f34d57eee80df3e6aa68c220c08df3e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
All children must be removed from a parent request before the parent is
freed.
Change-Id: I073ff0e9c5bcdd6181d90b918bfe4cce054f6c0b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Rename nvme_remove_child_request() to nvme_request_remove_child() and
move it next to nvme_request_add_child() to make the symmetry clear.
Change-Id: I78747c44ab3db1a656b33555a45f634dc5a55b31
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The NVMe specification recommends destroying all I/O submission and
completion queues before setting CC.SHN.
Change-Id: Iad71dd3fe03d897858034f3ca6ee02e0c55cc2b0
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The NVMe specification recommends that orderly shutdown should just
write CC.SHN while the controller is still enabled rather than writing
CC.EN = 0 first.
This also allows removal of the now-unused nvme_ctrlr_disable() and
nvme_ctrlr_wait_for_ready() functions.
Change-Id: I4702ffda153f218ebb8ed92f0e36144b7ceded93
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This can happen if the controller is still resetting as the SPDK NVMe
driver takes control.
Change-Id: I263ae8f2e7b271e0448450557452a115c90c4fb6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This patch is used to handle the memory leak issue when
a parent nvme_request is free. In our current code,
we did not free the nvme_request allocated by the children
in the exceptional case.
Change-Id: Iabd1f1c3594af60c38e74e3d96c14f78d1aa1aed
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
This patch is used to add a nvme_request remove child
helpler function
Change-Id: I1e5bb228d53333ca3601f4ae30fcd801ea39e532
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Drop the "data_transfered" variable and just update length, since length
was not used otherwise after this point in the loop.
Change-Id: Icd2991e4e85de7e8c951ba14c441434e871ea4ef
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
If only one descriptor is needed, there is a special case in the spec for
SGL1 using the Data Block descriptor type. Add a comment to make it
clear what is going on.
Also tweak the SGL1 setup to copy from the first SGL descriptor list
element instead of relying on the last value from the loop above, since
that could be easily broken by accident.
Change-Id: I49ef97fe5bf18d2bf1d86b4310a7d3abdfd03e57
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
tr->u.sgl is already a struct spdk_nvme_sgl_descriptor pointer.
Change-Id: Ie2c8c052fc28e6369d1d095b8d566acae47975d1
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
_nvme_qpair_build_hw_sgl_request() will only be called for payload_size
!= 0, so every SGL will have at least one segment. Drop the 'else' that
was handling nseg == 0, and add an assert to document the payload_size
requirement.
Change-Id: I48e2a862a7657ba85605c0d35c0b65dfac072167
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The assert is checking a variable that is invariant within the loop, so
move the assert up to the top of the function.
Change-Id: Iee7eea1736bc7f953665feb390c3d6340dbeffbc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This makes it easier to find the larger doc comments that produce separate
pages.
It also allows removing the lib/nvme directory from the Doxyfile, so
only the public API headers are used to generate documentation.
Change-Id: I8c46edb8067a91dda5b23fb0864efd3dd8aaeba5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
spdk_nvme_qpair_process_completions() is already documented in
spdk/nvme.h, so merge the doc comment from nvme_qpair.c into the public
header.
Change-Id: Id7722d99d209852ee64286e0a3fa127b863e10aa
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Don't allow the user to request more than the valid maximum number of
I/O queues (65535) or 0 I/O queues, since this can't be encoded.
Change-Id: I2d6e0bba03476085842bad683b273cdf9d6e6d5e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Per the NVMe spec, SGL segments must be Qword (8-byte) aligned. Add a
static assert to make sure this is true for the sgl member of struct
nvme_tracker (assuming the whole nvme_tracker is at least 8-byte aligned).
Change-Id: I827aa40b56de648d83f524a4f1e79c3202b676be
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
If the controller is failed, attempting to submit additional I/O is
futile - it will be immediately failed using the completion callback,
which can result in infinite recursion if the application code resubmits
I/Os on failure.
Instead, provide a way for request submission to indicate failure, and
use it to exit early if the controller is failed; this can only happen
when a reset failed (timed out).
If a request is submitted directly by the user when the controller has
failed, we can return an error code directly. For the case where I/O
was queued and is being resubmitted after a reset, we still need to call
the completion handler via _nvme_fail_request_ctrlr_failed().
Change-Id: I9e144328d524b25db2acf48e923b584746e8d0b6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Provide a new structure, spdk_nvme_ctrlr_opts, to let the user modify
the default controller initialization options during probe/attach.
Currently, only the number of queue pairs can be modified in this way;
other options will be added later.
Change-Id: Ie27b9429291d93a9353c0d820f0ad467d3b0e7cb
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Add a parameter to each I/OAT library function that requires a channel
instead of implicitly using the thread-local channel registration model.
I/OAT channels are already reported by the spdk_ioat_probe() attach
callback, so no infrastructure for channel allocation is necessary.
Change-Id: I8731126fcaea9fe2bafc41a3f75c969a100ef8f0
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Replace the previous code that allocated each tracker individually with
one large allocation per queue pair.
struct nvme_tracker is now explicitly padded to reach exactly 4096 bytes
to allow normal array indexing to work correctly while maintaining the
alignment requirement that ensures each tracker's PRP list does not
cross a page boundary.
This also allows removal of the act_tr array, since the tr array can be
indexed directly now, and each tracker can store its own active state.
Change-Id: Ia7c51735b96594d12f7f478cefcc4aedc84207ad
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The previous method for registering I/O queues did not allow the user
to specify queue priority for weighted round robin arbitration, and it
limited the application to one queue per controller per thread.
Change the API to require explicit allocation of each queue for each
controller using the new function spdk_nvme_ctrlr_alloc_io_qpair().
Each function that submits a command on an I/O queue now takes an
explicit qpair parameter rather than implicitly using the thread-local
queue.
This also allows the application to allocate different numbers of
threads per controller; previously, the number of queues was capped at
the smallest value supported by any attached controller.
Weighted round robin arbitration is not supported yet; additional
changes to the controller startup process are required to enable
alternate arbitration methods.
Change-Id: Ia33be1050a6953bc5a3cca9284aefcd95b01116e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Since the nvme_tracker struct was extended to allow space for 253 SGL
descriptors at 16 bytes each, we can use the same amount of space in the
other branch of the union to store 506 PRP list entries at 8 bytes each.
This increases the maximum supported I/O size for PRP-only devices from
128 KB to slightly under 2 MB.
Change-Id: I2b9905be41343ff360b4cdaccca87ea6f753e89c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Also add a compile-time assert to make sure this doesn't accidentally
break again in the future.
Change-Id: I4d18cfbf21392291e1bdd76eff055429009d28d6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
nvme_ctrlr_cmd_create_io_cq() was using the queue ID as the
IV (Interrupt Vector) field in the Create I/O Completion Queue command.
Since the SPDK NVMe driver does not enable interrupts, this is
misleading at best.
Change-Id: I3ea53701fdb9f21d9dc8d8fe20ccf2833b76cfbf
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This patch just implement the feature of format progress indicator.
the NVMe available does't support FPI currently.
Change-Id: Ie937591fb1720d8a062354322aabcc95ff14b2d3
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
This field is write-only in the current code; the NVMe library does
not track timeouts on requests.
Change-Id: I50e53bb3c299bf16912c48be8aad3eec829154af
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
When I/O spans a stripe boundary, the driver splits the request into
multiple requests, so for 1 segment memory larger than the stripe
size, we also need to split the segment memory.
Change-Id: I22ea5734d7066865a57a3c90fe18d5f76f373f1d
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
For those NVMe controllers which can support SGL feature in
firmware, we will use SGL for scattered payloads.
Change-Id: If688e6494ed62e8cba1d55fc6372c6e162cc09c3
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
nvme_spec.h already has a structure with the correct bitfields for the
CSTS register, so use it in struct spdk_nvme_registers.
Change-Id: Id0663aee2611fb5195f9012a3176799e32701bb0
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This will be exposed in the public API. This rename is in a separate
commit to ease review.
Change-Id: I1b7fef36f85265db27935ac4d22ceef3c7282502
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Set up the infrastructure for creating I/O submission queues with
variable queue priority (QPRIO in Create I/O SQ command).
Currently, this is unused, since we always use the default arbitration
method (round robin), but it will allow reinitializing submission queues
with the correct priority once weighted round robin is supported.
Change-Id: I425003879e624cfcc9687bdc495b5c1726b5a8af
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This function no longer exists and was not part of the public API.
Change-Id: I94fd066b63e812367687d11bc00aa11ab88d4671
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Many of the internal controller initialization functions did not check
for allocation failure; add return codes and check them where
applicable.
Change-Id: Id1b33bb06fca84035369d8b7ecd4c36b8ba7134c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This function returns true if the namespace is active or false if it is
inactive (e.g. no namespace has been attached to the specified namespace
ID yet).
Also use the new function to add checks in the examples and tests where
applicable.
Change-Id: I35465b315ae1a1677c5a82191ad9b1da1c216d50
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Prepare for qpair to be exposed as part of the public API.
Change-Id: Ia63e863e95554adceeade20c829f12fe346375d5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
When multiple NVMe controllers are being initialized during
spdk_nvme_probe(), we can overlap the hardware resets of all controllers
to improve startup time.
Rewrite the initialization sequence as a polling function,
nvme_ctrlr_process_init(), that maintains a per-controller state machine
to determine which initialization step is underway. Each step also has
a timeout to ensure the process will terminate if the hardware is hung.
Currently, only the hardware reset (toggling of CC.EN and waiting for
CSTS.RDY) is done in parallel; the rest of initialization is done
sequentially in nvme_ctrlr_start() as before. These steps could also be
parallelized in a similar framework if measurements indicate that they
take a significant amount of time.
Change-Id: I02ce5863f1b5c13ad65ccd8be571085528d98bd5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This check was dead code, since both places that called
nvme_ctrlr_wait_for_ready() could only ever have cc.en = 1.
Remove the original nvme_ctrlr_wait_for_ready() wrapper and rename
_nvme_ctrlr_wait_for_ready() without the underscore to replace it.
Change-Id: I6c9aa6a5b93606fb89d168c23f6735fcf3a84eaa
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
In nvme_ctrlr_hw_reset(), if we encounter a controller whose CC.EN bit
is already 0 (controller is disabled), the previous code would enable
the controller just so that it could be disabled to get a full reset
(transition from CC.EN = 1 to CC.EN = 0). However, it is a safe
assumption that if CC.EN is already 0, the controller has just been
reset, so we don't need to reset it again.
This saves a significant amount of time (2+ seconds per controller with
Intel SSD DC P3700) during initialization for devices that were disabled
on startup.
Change-Id: I552b1f0f185a84a8a0ce57a93b012d9d5fe096f3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
cleanup.sh and unbind.sh have been combined into a single
setup.sh that takes one optional parameter (reset). If no
parameter is given, the script will automatically bind
all NVMe and IOAT devices to either uio_pci_generic
or vfio-pci, as appropriate based on IOMMU settings. If
the reset parameter is given, the devices will be bound back
to the appropriate kernel drivers.
Change-Id: I25db3234f1ecfb352a281e5093f4c1aa455152ae
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Explicitly include system headers for types that are used in public
headers.
These were being pulled in by example code, so SPDK itself would build,
but other apps that did not include stdbool.h would fail to compile when
including spdk/nvme.h.
Also include nvme.h first in nvme_internal.h so this case gets tested
during normal compilation.
Change-Id: I8ed0fc4e0dcf71551738c461b4b825cc2ee1d233
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
payload here is a pointer to the buffer, not a struct nvme_payload.
Use nvme_allocate_request_contig() and pass the length in bytes rather
than dwords.
Change-Id: Idbbb3614b1d69148fe041d26e0c148bd9ce53724
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The parent field is in the cache line of nvme_request that is only
supposed to be accessed for split (child) I/Os. All accesses to parent
are done from child-specific calls now, so it does not need to be
initialized in the common case of a non-split I/O.
nvme_request_add_child() will set parent when splitting occurs.
Change-Id: Ib86c16ba1ea2ce32f62079831101da2a099047af
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Allows simplification of _nvme_qpair_build_sgl_request(), which does not
need to know whether a request is a child or not.
This also removes a read of req->parent for non-split I/Os; the parent
field is in the section of nvme_request that is only intended to be
initialized for split I/Os, which should be detected by looking at
num_children.
Additionally, this fixes a potential problem if requests were nested
more than one level deep (e.g. req->parent was not the original user
request).
Change-Id: I3ea1dc134bbb1e3b8c6b5a479f5d760bd97ea848
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This patch fixes the following issues in ioat_impl.h
while CONFIG_PCIACCESS is set to n:
1 lack of the definition of ioat_pcicfg_unmap_bar.
2 Define a macro to remove the duplicated vendor code
for ioat_driver_id enumration.
Change-Id: I4011098ac296d1ec320bffb5ffa6e098b70a5545
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Building the tip of the spdk master against the dpdk-2.2.0 fails with
inappropriate RTE_CACHE_LINE_SIZE error. The simple reversal of the RTE
include file order below fixed it for me.
Change-Id: I8782b7ee21d7f185e6e678f874fbdab9403117a5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This patch is used to remove the duplicated code.
As we found the structure in "ifdef and else" are same.
Change-Id: I1717ce3dcc14134ac59c165d801e5e811b987be5
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
This patch is used to support spdk_pci_device_cfg read/write
with byte and word size.
Change-Id: I49084e231bd6b5f5b22180a3eb36ddad4430b3a4
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Move toward collecting PCI IDs, class codes, etc. in pci_ids.h instead
of individual device-specific headers.
Change-Id: Icff162d48ac663db71d0576ceee16a9bd7a751cd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
PCI_VENDOR_ID_INTEL -> SPDK_PCI_VID_INTEL
Also change the inclusion guard macro to be consistent with the other
SPDK headers.
Change-Id: I29346267172cb8c07cc4289eed4eca2d55e942d6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This doesn't need to be part of the public API. It is only used by the
NVMe quirk lookup tables.
Change-Id: I7662e333c70b7c5f814bd7c8a528b6bff1f0732e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Rename all functions with a spdk_ prefix, and provide enough of an API
to avoid apps needing to #include <pciaccess.h>.
The opaque type used in the public API for a PCI device is now
struct spdk_pci_device *.
Change-Id: I1e7a09bbc5328c624bec8cf5c8a69ab0ea8e8254
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is a step toward abstracting PCI access so that libpciaccess can be
swapped out more easily.
Change-Id: I5491459460cbfbd0be471f70f9d07a7eb3175234
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Instead of writing the completion doorbell once per completion,
just write it once at the end of the completion while loop.
This reduces the number of mmio writes by coalescing several
writes into one when we get multiple completions at a time.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I3cc3864dcfe43186bec51be1a732e84ef3be05ae
Similar to the NVMe API change, this allows better abstraction of the
PCI subsystem.
Change-Id: I2b84d9c3c498a08d4451b4ff27d0865f0456c210
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This patch did the work to use pci related functions
provided by DPDK.
Change-Id: I263b79f1b42868ef0c1efcf1bc392a4b3a328e93
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
With CONFIG_PCIACCESS=y in CONFIG file, we can
use libpciaccess library; With CONFIG_PCIACCESS=n
in CONFIG file, we use pciaccess functions provided
in DPDK.
Change-Id: I786c5589b8e7909ba2e59d222938dd5ba45bf92d
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
The new probing API will find all NVMe devices on the system and ask the
caller whether to attach to each one. The caller will then receive a
callback once each controller has finished initializing and has been
attached to the driver.
This will enable cleanup of the PCI abstraction layer (enabling us to
use DPDK PCI functionality) as well as allowing future work on parallel
NVMe controller startup and PCIe hotplug support.
Change-Id: I3cdde7bfab0bc0bea1993dd549b9b0e8d36db9be
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
For those Crystal Beach DMA channels which support block fill capability,
we add a fill API here that can zero out pages or fill them with a
fixed pattern.
Change-Id: I8a57337702b951c703d494004b111f6d206279fb
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
This patch adds Intel NVMe device list and overrides the
supported log pages according to the quirk list.
In particular, the READ_CMD_LATENCY and WRITE_CMD_LATENCY pages are
supported on Intel DC P3x00 devices despite not being listed in the
Intel vendor-specific log page directory.
Change-Id: I3a2b6a5fa142c6e9c93567df65e85980bd3c7cc0
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
Also add a space between Copyright and (c).
The copyright year can be determined using git metadata.
Also remove the duplicated "All rights reserved." - every instance of
this line already has a corresponding "All rights reserved" immediately
below it, except for examples/ioat/kperf/kmod/dma_perf.c, where I have
added it manually.
Performed using this command:
git ls-files | xargs sed -i -e 's/Copyright(c) \(.*\) Intel Corporation. All rights reserved./Copyright (c) Intel Corporation./'
Change-Id: I3779f404966800709024eb1eb66a50068af2716c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
1 Add new API nvme_ctrlr_is_feature_supported().
2 Add unit test for new API.
Change-Id: Ia6d8710755c3b13984fca9d56700efe043be1402
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
This will allow replacing these _nvme_fail_request_bad_vtophys() calls
with the correct error later. vtophys is not actually used within the
SGL request builder, so this is the wrong error.
Change-Id: Ibc2a3b029a8abad1d563b9df200325d7d64498da
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
No code change, just moved into a function for readability.
Change-Id: I883443c06d961c6dbeffed1a6fb153177e6e3fcd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This cleans up the I/O splitting code somewhat.
It also moves the SGL payload function pointers up into the hot cache
section of struct nvme_request without pushing the other important
members past the cacheline boundary (because payload is now a union).
Change-Id: I14a5c24f579d57bb84d845147d03aa53bb4bb209
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
A namespace indicates support for reservations by reporting a non-zero
value in the Reservation Capabilities field in the Identify Namespace
data structure, and controller indicates support for reservation in the
Identify Controller data structure, Here we used namespace field as the
support flag.
Change-Id: I0e1e29548aa3fc8b6d3bbeb4149ec4864316f092
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Support for the Force Unit Access and Limited Retry
bits on reads and writes.
Change-Id: I9860848358377d63a967a4ba6ee9c061faf284d4
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This patch is used to wrap the inner implementation
of libpciaccess and prepare for the same interface
to applications in the future patch
Change-Id: I4d40fae0bd86b451ed38dbfd9bcc015f9bfc8436
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
NVMe reservations provide capabilities that may be used by two or more
hosts to coordinate access to a shared namespace, here we add the 4
reservation commands: reservation register/acquire/release/report.
Change-Id: Ib03ae2120a57dd14aa64311a6ffeb39fda73018c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
For the purpose to support different types of input scattered payloads,
such as iovs or scattered list, we define common method in the NVMe
driver, users should implement their own functions to iterate each
segment memory.
Change-Id: Id2765747296a66997518281af0db04888ffc4b53
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Change nvme_ctrlr_is_log_page_supported() to match
nvme_ctrlr_cmd_get_log_page().
Change-Id: I4c8a1f11044b083f8f8990ef40a4f789fa3c24e3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Request allocation may fail, so we need a way to indicate failure to the
caller.
Change-Id: I278c3f42e4d2fa1902bb0ab33ad3bf7c7007fd0d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
1 Add supported log pages data structure.
2 Bulid up supported log pages when NVME start.
3 Provide unified API for getting log pages.
3 Unit test suit optimization base on above modification.
Change-Id: I03cdb93f5c94e6897510d7f19bc7d9f4e70f9222
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
Use /sys/bus/pci/devices/.../driver to determine which driver is loaded
for a particular device.
Change-Id: I5859a776e524033e1c6d6ec3796b7e11bdcf0bc4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This more accurately represents what function it performs.
Also remove pci_device_has_uio_driver() from the public API. Callers
should use pci_device_has_non_uio_driver() instead.
Change-Id: I9623fe1345b43e981d5823804e33d01ac0d3bb1c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
nvme_ctrlr_process_io_completions() and
nvme_ctrlr_process_admin_completions() now return the number of
completions processed.
This also adds the possibility of returning an error from the
process_*_completions functions (currently unused, but this at least
gets the API ready in case error conditions are added later).
Change-Id: I1b32ee4f2f3c1c474d646fa2d6b8b7bbb769785f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This ensures that any uninitialized fields are 0/NULL so if
ioat_channel_start() fails, ioat_channel_destruct() will not try to free
bogus pointers.
Change-Id: I99278c9fa280cbcdf3f7448e77db3ac98b59cdd6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Previously, if nvme_allocate_request() failed in
nvme_ctrlr_construct_and_submit_aer(), there was no error checking, so a
NULL pointer would be dereferenced.
Add a return value to nvme_ctrlr_construct_and_submit_aer() so we can
signal failure to the caller. This can only really be reasonably
handled during initialization; when resubmitting a completed AER later,
there is nowhere to report failure, so the AER will just remain
unsubmitted.
Change-Id: I413eb6c21be01cd9a61e67f62f2d0b7170eabaa3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The ioat library currently only supports DMA copy operations, but the
hardware can do other types of transfers. Add a union of the hardware
descriptor structures to enable support for the other operations in the
future.
Also add a generic hardware descriptor type to allow access to the parts
of the descriptor that are common between all types.
Change-Id: I3b54421ce771f58b78910e790b53026f311f918e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
According to the specification, the dataset management for deallocate
attribute can support to 256 ranges, so we should use uint16_t
instead of uint8_t as the ranges parameter.
Change-Id: Ibacc00da8b4b9e2b2f3454d382aadf7ad353ff31
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The parent field is no longer used in the normal (non-split) I/O path,
so move it down to the default-uninitialized part of struct nvme_request
that is only touched for parent/child I/O.
This also puts it closer to other related fields (children,
child_tailq, parent_status) for improved readability.
Change-Id: I120df1df0c967d2f74daa6e97c0bc83626e3be7f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
nvme_qpair_submit_tracker() and nvme_qpair_manual_complete_request() are
only used from within nvme_qpair.c, so they can be static.
nvme_qpair_submit_tracker() is moved up to avoid needing a declaration
(no other code change).
nvme_ctrlr_hw_reset() is only used from within nvme_ctrlr.c, so it can
be static.
Change-Id: I9a7953d7baaec76e875dd535daf557ea24bef801
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These delays are left over from early development. They are completely
unnecessary and not based on anything in the NVMe spec.
Startup time should be slightly improved (on the order of 100 ms in
normal cases).
Change-Id: I9068b1a0f42feabcfe656d68be91e05a56cc53a3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This lets us signal an error if the channel is halted in
ioat_process_channel_events().
Change-Id: Iffaf4fd1e27d1254f9d95a37d732ae4a5f3a0465
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Rather than individually allocating each ring entry, use two large
allocations, one for the hardware descriptors and one for the software
descriptor contexts.
This allows the use of simple array indexing on the rings and also
allows the removal of most of the software descriptor structure,
since the necessary information can be retrieved based on the ring
index now.
Change-Id: I73ef24450f69ca0fc35e350286282c6b1c77a207
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
NVMe doesn't require the specific 64-bit MMIO ordering on 32-bit
platforms performed in spdk_mmio_read_8(), but it doesn't hurt.
We have to pick one of the two possible orderings, so pick the one
required by I/OAT.
Change-Id: I2b909d64d0c077b797d0f64a11d78d1ecc55eec7
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The ioat driver supports DMA engine copy offload hardware available on
Intel Xeon platforms.
Change-Id: Ida0b17b25816576948ddb1b0443587e0f09574d4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Use $(CLEAN_C) throughout the Makefiles to clean up a consistent set of
generated files.
This also adds coverage files to the list of cleaned files.
Change-Id: Iceb922935a45c9eecbf2f3443bd0ee4f5c966825
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Most devices today support far fewer than 1024, but this is a
more reasonable default upper limit than the spec-defined 64K.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia8a6d80c3a5aa181f27c8354758c6ca468013d92
lib/memory was already using this pattern; extend it to lib/util and
lib/nvme.
Change-Id: I84a633d7805522fc94d8fc11ad5486ce552702e5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The .o files are always kept anyway, so there is no need for an explicit
rule.
Change-Id: Id1687ba89daabfda5802e4328deb127403277928
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Add common $(COMPILE_C), $(LINK_C), and $(LIB_C) variables that contain
the commands to build a .o from a .c, an app from objects and libraries,
and a library from objects, respectively.
Change-Id: Ie2eaa13156b8bf3db7a4ffa66161382d829aef07
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
nvme_ctrlr_process_io_completions() now takes a second parameter,
max_completions, to let the user limit the number of I/Os completed on
each poll.
If there are many I/Os waiting to be completed, the
nvme_ctrlr_process_io_completions() function could run for a long time
before returning control to the user, so the max_completions parameter
lets the user have more control of latency.
Change-Id: I3173059d94ec1cc5dbb636fc0ffd3dc09f3bfe4b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
After converting is_resetting to bool, it is smaller and can be packed
more efficiently with is_failed and reordered after the larger fields
used in the I/O path.
Change-Id: Ifa2301eb61ce8d38eb5412cca61d2a91b1474101
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
It was previously uint32_t because it was accessed with special
uint32_t-only atomic read/write helper functions, but that was replaced
with normal variable accesses protected by a mutex.
Change-Id: I304a7ef8c723cb33fd08110b697f848823a163e7
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Set SPDK_ROOT_DIR explicitly in each Makefile so that make from a
subdirectory will work (assuming all dependencies from the upper
directory have already been built). This allows partial rebuilds of the
source tree, as well as building the unit tests without requiring DPDK.
Change-Id: I3f65b805d490b40ff5ec53cceb61df542ce814f1
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This helps weed out functions that should be static, functions that are
not declared in public header files, and .c files that don't include
their .h interface headers.
Change-Id: Ie39f83ad4b320847e4a938bd1d4d0b4fa21c2ffa
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Fix all of the uses of __thread so they are at the beginning (similar to
e.g. static).
Don't actually enable -Wold-style-declaration, since clang doesn't
understand that.
Change-Id: I0dcbb758143eab90fc978334c8f256c6602cc4cd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The ioctl() calls in dev_get_blocklen() were checking for != 0 instead
of == 0, so the default path (512) was always being taken.
Change-Id: Ib0b016b1d453fb94d408063417b7485ff24ed220
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Rename the nvme_free_request macro to nvme_dealloc_request to match
nvme_alloc_request and add a wrapper function to nvme.c so that the
macro contents are only expanded once.
The DPDK nvme_impl.h uses rte_mempool_put(), which generates a large
amount of code inline. Moving this macro expansion to a wrapper
function avoids inlining it in the multiple places nvme_free_request()
gets called, most of which are error handling cases that are not in the
hot I/O path.
Change-Id: I64ea9c39ba47e26672eee8d5058f1489e07eee5b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Like sprintf() with automatic buffer allocation.
This should help to avoid fixed-size buffers in
non-performance-sensitive code that formats strings.
Change-Id: I35209ae84014ed5daf41baa5b03af8a5f6b02b8e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is the only place nvme_request_add_child() is used, so move it
nearby and make it static to allow the compiler to inline it.
Change-Id: If4a7e17fde0b0272e1d4432c1dcedbec27c25371
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Move dependency includes into a new spdk.deps.mk file,
then include it at the end of Makefiles that build
source files.
Also add a test to autobuild.sh to confirm that
binaries are regenerated if we make after touching a
header file.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: If6a1905706a840f92cbdf3ace7fbdb27fe2de213
Pull the almost-identical request splitting code for driver-assisted
striping and maximum I/O size into its own function,
_nvme_ns_cmd_split_request().
Change-Id: I3c15ac2073f8f5aec721c427199c8fb1a5d6a1fc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This helps enable FreeBSD, where pciaccess pci_device_has_kernel_driver()
is not functional. The function will return 0 if there is no driver
attached, or the Linux uio or FreeBSD nic_uio driver is attached. It will
return 1 otherwise.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I0921e61c9040b1e0411b5dc40b36fc7f2721c8c5
The changes are minor:
- remove unneeded error.h
- replace PATH_MAX with a suitable local #define
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I5aecf8b53e0ac7582f394c71b4668888a6c6292f
The Linux pagemap-based implementation obviously does not
work on FreeBSD. DPDK has data structures describing the huge
pages it has allocated, so use that instead when we need to
populate new 2MB mappings in our tables.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I924e104f42891aaa2f931159aabba2779f239e91
GCC generates a series of 64-bit MOV instructions for the memcpy() into
the submission queue. We can do better with 128-bit SSE2 instructions.
DPDK already has a memcpy implementation that is optimized for small
inline copies, so use it instead of memcpy.
Change-Id: I5f09259b4d5cb089ace4a8ea6d2078c03fee84f3
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
No change in behavior, just a simplification.
We already have a check for retry, so pull the cb_fn check out and put
it under the !retry branch.
This makes it clearer that requests that are going to be retried will
not get their callbacks called.
Change-Id: I70c7067e550c7fca78b0441b5474833f73863315
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Since it is empty function. It can be deleted.
Add nvme_assert to fix issue reported by scan-build.
Change-Id: Ia0e8f656e1dac0da7ec72f8404469ea1b0dcb40e
Signed-off-by: Liang Yan <liangx.yan@intel.com>
This is the only place that was using printf directly in the NVMe
library. Replace it with the official nvme_printf logging mechanism.
Change-Id: I689a7c0854b5e47eb357150f814e347cd44be79c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
In nvme_qpair_complete_tracker, make sure we got a valid request in the
tracker that is being completed.
This should never occur in practice, but safeguard against it in case of
programmer error. Fixes a scan-build warning about potential NULL
dereference.
Change-Id: Id82af604d2a5ed5de0aeccf3affa1900f6712ebe
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Document the assumptions made by qpair_construct using asserts.
These values can't actually be 0 in practice due to the way they are
derived, but scan-build can't see that. It is also useful to have these
asserts in case of future modifications.
Change-Id: I546c057f5cbe7ccc62acd90b595e423cd450d86a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
nvme_dump_command is totally unused aside from the unit test.
nvme_dump_completion was used in qpair, but it can be replaced with the
equivalent nvme_qpair_print_completion.
Also added the missing nvme_completion fields to nvme_qpair_print_completion
that had been printed by nvme_dump_command.
Change-Id: Ia5ee66f3553df06febe8f465d42e49a84c555dd2
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This is calculated elsewhere now, so remove the comments around
nvme_qpair_construct calls.
Change-Id: I2dc4956a9e250b88e62038bc55cdd315940ad391
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
rc is reinitialized before it is ever read.
Change-Id: I9abbc256fb06022f3024b0aa3827be02a273f20a
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
