When we do frequent same subsystem add/delete,
we will face the adding issue. For example,
1 Add subsystem A
2 Delete subsystem A
3 Add subsystem A (Fail in this step).
The reason is that we did not correctly free
the listener resources of subsystems, and this patch
can solve this issue.
Change-Id: I6765a306a3f10c9a0f38c95dbba12e2a4073e705
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
They were very close to the same already, so finish the job.
Change-Id: Ifba9e3b2d11a3e70cbfbe46f57a67552db2757ed
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The probe_info was reduced to just containing a
transport_id, so remove probe_info entirely.
Change-Id: Ica9a22d126cd14e282decd3eea1a0afe0460f099
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Added the socket ID to the NVMe over fabrics target warning messages.
This will help during troubleshooting to check that the NVMe device, Subsystem and NIC are on the same NUMA node
Change-Id: Id750a5a7694c898bb61ea26089ea3098611d530a
Signed-off-by: John Kariuki <John.K.Kariuki@intel.com>
If AcceptorPollRate is configured in configuration
file, the default value used will be given by
ACCEPT_TIMEOUT_US. So change the default value, it
can solve the performance degradation issue of
nvmf target.
Change-Id: I867bb03dd8b2b81b86911130babd0334d9857de8
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Remove usage of the conf structs so they can be moved out of the public
API header.
Change-Id: I1c7375ec7708b323f50af09aeb7b2b2c9c770df4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The details of the structure were removed earlier, but
now remove all references even to a pointer to the
structure. The user can refer to transports by their
string name.
Change-Id: I273356f46329ea5372dcd951eda6f14767477d69
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This is a step toward abstracting away the definition
of the subsystem.
Change-Id: I88b2aa107b27152620f51a1ca2a153792b4c85e9
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
spdk_nvme_probe() will now provide a struct spdk_nvme_probe_info to the
probe and attach callbacks in place of the PCI device pointer.
This struct contains the useful information that could be retrieved from
the PCI device during probe.
The goal of this change is to allow expansion of the probe information
in the future when other transports (specifically, NVMe over Fabrics)
are added that do not necessarily use PCI addressing or device IDs.
Change-Id: I59a2a9e874e248ce5fa1d7f4b57c8056962ff3cd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Add a helper function that converts a PCI address from a string into a
struct spdk_pci_addr and use it in place of the various sscanf()
invocations throughout SPDK.
Change-Id: Id2749723f76db741567e01b4bcb0fffb0e425fcd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The NVMe over Fabrics target was storing the PCI device pointer for each
direct-mode controller, but it only really needs the PCI address, which
is exposed via the get_nvmf_subsystems RPC.
Also update the same code path to use the new spdk_pci_device_get_addr()
function for brevity.
Change-Id: I0708b3331b7c279c1a86f0d7459b5deb40dd7c89
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This changes as little code as possible while still creating
a single public API header. This enables future clean up
of the public API and clarification of the exposed
concepts.
Change-Id: I780e7a5a9afd27acf0276516bd71b896ad301c50
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Starting the subsystem consists of firing an event
to the subsystem poller's lcore and performing
I/O channel initialization (for virtual NVMe
subsystems) before starting the poller.
Previously the subsystem would be started immediately
when created, which was before the subsystem's
mode and other parameters (such as a virtual subsystem's
bdevs) have been set. This resulted in no I/O channels
being allocated in virtual subsystem mode.
So break out the start code into a new
nvmf_tgt_start_subsystem() function, which clients
must call after fully initializing a subsystem created
with nvmf_tgt_create_subsystem().
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I56534668887ef99d2e892844acd12194920c3245
Extend the check added in commit
7552ee5512f6a6eccbdfbde8554ddd2fc1f877a4 ("nvmf_tgt: fail if NVMe
controller is not found") to the RPC interface.
Also remove the default PCI address from rpc.py, since it is not a
useful default - the user must specify a PCI address, which will vary
depending on the configuration of the machine.
Change-Id: If9a4e19ef2a84c9d839b467abc011046613f7168
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Instead of a magical "All" value for allowing all hosts, allow the
caller of construct_nvmf_subsystem to omit the "hosts" parameter.
Change-Id: I97c7ae806a0be7142fb59708d47023e42e127fcc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
create_nvmf_subsystem should not allow the user to specify any
namespaces in direct mode.
Also remove the default Malloc0 namespace in rpc.py; the user
should be required to specify the list in virtual mode.
Change-Id: I3aaf7c64b9ceee89a3f7db15760a669527adb29d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
'virtual' is a keyword in C++, so avoid using it in variable
and structure names in case any files are eventually
included from a C++ project.
Change-Id: I2122750445def63038af68a3000758e33b937f9d
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
These 4 callbacks can be condensed into two callbacks, which
simplifies the API.
Change-Id: I069da00de34b252753cdc8961439e13a75d1cc68
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This patch also drops support for automatically unbinding
devices from the kernel - run scripts/setup.sh first.
Our generic pci interface is now hidden behind include/spdk/pci.h
and implemented in lib/util/pci.c. We no longer wrap the calls
in nvme_impl.h or ioat_impl.h. The implementation now only uses
DPDK and the libpciaccess dependency has been removed. If using
a version of DPDK earlier than 16.07, enumerating devices
by class code isn't available and only Intel SSDs will be
discovered. DPDK 16.07 adds enumeration by class code and all
NVMe devices will be correctly discovered.
Change-Id: I0e8bac36b5ca57df604a2b310c47342c67dc9f3c
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
For NUMA architecture, you may get performance drop if your NVMe
devices and transport devices located in different socket, with
this patch, the NVMf target will print a warning message to remind
the user that your configuration is not optimal.
Change-Id: Ia6013ef95984f0ce8c7f1ca86b89c0375686a188
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
1 In our nvmf tgt implemention, we use the async
mode to delete the nvmf subsystem. However, when
we parse nvmf subsystem, we need to use the sync
function to delete the nvmf subsystem. Since if
there is error, we will call spdk_app_stop, thus
async functions will not be executed. It is
approved in my local test.
2 Add debug info in spdk_nvmf_delete_subsystem
Change-Id: Ia8ecd6eee1bbd25cb3e1ceeb0e2146f3f03be228
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
For subsystems in Direct mode, make sure the NVMe controller specified
is actually located during the probe call.
Change-Id: I7d5212ea628eadba3e720172a5ead8902eaaf6d0
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The application is now entirely responsible for scheduling subsystem
pollers and sending events between threads.
Change-Id: I88da1f53b5e8852c7c4acd6f0a7a1e2219fbed41
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Need to check bdev != NULL, otherwise there will be a segment
fault later
Change-Id: I8aa74e7716decc3166f9f2abbbb0c8ff73816f84
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Match the direct NVMe device backend by printing out each block device
that is added as a namespace to a virtual subsystem.
Change-Id: I5c2762ca1fbab3529c19c822c95c73b72e9dabbc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Move the NQN validation into the subsytem creation function, and fix the
allowed size to match the spec.
The spec is not clear about the allowed NQN size; for now, interpret it
as 223 bytes, including the null terminator (222 bytes of actual NQN
plus one terminator byte).
Change-Id: If9743ab2fe009d9d852e8b03317d9b38d8af18dc
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This header will be used as the general NVMe over Fabrics target app
include file.
Change-Id: Ia26ff6d97fb3fd3f2c55b43c15abbb0e58998e30
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The NVMe over Fabrics target library now exposes a simple function call
that polls the acceptor once, and the application handles registration
of the poller.
Also rename the transport function pointers related to the acceptor so
they better reflect their purpose.
Change-Id: I5fa0d516586bf17e73afeb88ff3c2d5b0d46794d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
