In order to populate our PCI device list with devices
located behind the VMD, we'll need to fill out those
device structures from within a special VMD driver. That
driver will base on PCI configuration and BAR accesses,
but definitely not on DPDK. We want to put the VMD driver
outside of the env lib, so we provide it with a direct
access to the device struct.
Change-Id: Iabddf361a805e69d7e857c2d07ceaed36aca261d
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/435800
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
In order to populate our PCI device list with devices
located behind the VMD, we'll need to fill out those
device structures from within a special VMD driver. That
driver will base on PCI configuration and BAR accesses,
but definitely not on DPDK. We want to put the VMD driver
outside of the env lib, so we're about to provide it with
a direct access to the device struct. Before we do that,
let's group all the env-internal fields into an extra
struct "internal".
The spdk_pci_device struct does actually depend on DPDK
now as it contains an `rte_pci_device *dev_handle` field,
but we can easily break that dependency. The field is only
used as an arguement to DPDK functions, so we can change
its type to void* and let the implicit type conversion do
the magic. After all, the VMD driver will potentially use
it to store its (non-DPDK) data as well.
Change-Id: I425d6dfa7af13e022f5377ceaff39efbd4a01b3d
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/435799
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
libpmem only allows passing a size when CREATE flag
is set.
This requires some updates in the unit test stubs
for pmem_map_file as well. While here, do some
additional cleanup and add a g_volatile_pm_buf_len to
track the size of the allocated volatile pm buffer.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ib9fe58fd9946161dd20bb8391be2e9680705ab22
Reviewed-on: https://review.gerrithub.io/435945
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
When the caller of the RPC API has all the necessary information about
how to access a Ceph cluster, then having to create configuration
files before calling the RPC API is problematic (has to touch files
owned by a local admin, changes must be removed again).
But having to encode support for certain configuration options in SPDK
is also problematic, because that might change depending on the
librados version.
The approach taken here is to merely pass through arbitrary key/value
config options. Existing config files are ignored when that happens.
The caller of the RPC then has full control over the connection setup
and can be sure that he does not inherit settings from a local file
accidentally.
In addition, user management is supported now, with or without a
config. This is useful for accessing a volume with a less privileged
user. Previously, passing NULL to rados_create implicitly chose the
"admin" user.
Change-Id: I5e7f36092df663a3d7ac503c04fc624a8fe1208e
Signed-off-by: Patrick Ohly <patrick.ohly@intel.com>
Reviewed-on: https://review.gerrithub.io/430460
Reviewed-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
SCSI persistent reservation feature need to get the TransportID
for the I_T nexus, so when creating initiator port we also
set the TransportID according to the specification, while here,
we use the format code 0x1 for the TransportID.
Change-Id: Ib45bec04bf0e33e2b0f611dd3846597f4176d069
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/435212
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Ensure that no trace point group IDs are ever duplicated.
Arrange trace registration in order on tgroup_id.
Change-Id: Id72600257780b1ab95b25c85daaa78c392a9479f
Signed-off-by: Liu Xiaodong <xiaodong.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/435571
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Since we use aligned buffer, I think that the error handling
path here is not correct, the address is wrong.
Change-Id: I5bcb7f050199496423f861fd6aea65e0fe48c804
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/435992
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
spdk_rpc_is_method_allowed() allows to check if method is permitted for
given state.
Change-Id: I0b0046482262dfc7fa521647991eb88a38e4c1d3
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/430487
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
When the applications call spdk_nvme_ctrlr_alloc_io_qpair,
there will be cmd to the admin qpairs in nvme_ctrlr_get_cc,
so there is contention. We should use the lock to protect
nvme_ctrl_get_cc. Otherwise, the multiple threads will have
contention on the admin qpair, thus there will be coredump issue.
We get the bug when testing NVMe-oF TCP transport, and this
patch can address this issue.
Change-Id: I7247f98cdf890c2eafaf8fb94580ecd714010bd5
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/435577
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
DPDK 18.11+ multi-process hotplug isn't robust.
Multiple secondary processes starting at the same
time might cause the internal IPC to misbehave.
Just retry hotplugging/hotremoving the device
in such case.
Change-Id: I1f830c2c0dbe1d63eca9a116101b3d202172b2ca
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434539
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
With all the error checks and segfault preventions in place,
we can finally enable hotplug in a multi-process scenario
for DPDK 18.11+.
If a device is attached in the primary process, it will send
an attach IPC request to the secondary process which needs
to succeed. Until now it would get rejected, and the attach
would fail in all the processes.
The device in secondary process will be now probed by DPDK
and will be put into the process local SPDK list of devices
to be locally attached. Either SPDK will attach it sometime
later on any attach/enumerate request, or DPDK will remove
it automatically once the same device in the primary process
gets removed.
We also allow the surprise attach in primary processes, as
it's technically possible for the pci devices (NVMe) to
be attached exclusively from the secondary process. The
fact that the NVMe stack doesn't support it is another story.
Currently the NVMe stack will handle the failure by itself
just fine.
Change-Id: Ia24a8b4610cc7c659f59a2fdda9d8a78e58af873
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434416
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
DPDK 18.11+ does its best to ensure all devices are
equally attached or detached in all processes within
a shared memory group. For SPDK it means that if
a device is hotplugged in the primary, then DPDK will
automatically send an IPC hotplug request to all other
processes. Those other processes may not have the same
SPDK PCI driver registered and may fail to attach the
device. DPDK will send back the failure status and the
primary process will also fail to hotplug its device.
To prevent that, we need to pre-register the pci
drivers on env init.
We register the drivers just after the EAL init
because we don't want the matching devices to be picked
up by the initial bus probe in DPDK. That's for 2 reasons:
1) we don't want to attach *all* available devices
2) devices attached from non-SPDK context (that is,
outside of the spdk attach or enumerate functions)
will still fail to attach - the entire attaching
process will only take significant amount of time
and will bloat the log with useless status messages
Change-Id: I7b4c3a2e355f98ea755649f789137f5a727bc935
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434415
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Although the struct is used as an enumeration context,
it really is a pci driver. The subsuequent patch introduces
a few functions around the pci driver, so rename the struct
to make it align nicely with those functions.
Change-Id: I919c30e55d9f42d795ecd8e20e5d29f3918c17a5
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434414
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Upon detaching a device in a secondary process, DPDK 18.11
will try to detach it from the primary process as well.
SPDK doesn't support such hot-detach and will reject it
in the primary process. That will cause the secondary
process to also reject its detach. The device in the
secondary process will be still there in DPDK, but for
SPDK it will remain inaccessible - neither attach, nor
enumerate will work on it.
To fix it, we make our attach and enumerate functions
always check the process local list of devices probed
by DPDK, but not attached in SPDK.
Looking at the patch from a different perspective, it
simply introduces error handling for the DPDK detach
function. If a device failed to detach, we'll now maintain
it locally in SPDK to make it attach-able again.
Change-Id: I8c509a571bea7a9fb413c9c2bfd64c62ad91074b
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434413
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
It's handy to store the SPDK structs within the device
structure. The subsequent patch will make us use
spdk_pci_addr much more frequently, so it makes sense
to keep it around rather than build it up from rte_pci_addr
everytime.
The upcoming VMD driver will also benefit from this patch
by being able to fill the spdk_pci_device struct with any
custom PCI details.
Change-Id: I236a19e28beba9a593b29f23b79b1b0b92ef1fa7
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434418
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
In DPDK 18.11, a device can be potentially detached not only
upon an SPDK request, but also directly from within the DPDK
itself. In a multi-process scenario, when one process detaches
the PCI device, an IPC message - detach request - will be sent
to every other process in the same shared memory group. As we
don't propagate the removal notification to upper layers, the
still-referenced rte_pci_device object will just disappear at
one moment.
SPDK is still not ready for supporting the above case and will
try to avoid it, but just in case some detach request slips
through, then this patch provides the sanity checks preventing
SPDK from crashing.
Change-Id: I3e35d8efb33085163b9acd8a565e86a4221df844
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434412
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Very minor cleanup before we start refactoring the code.
Change-Id: I00d768ec0c84f2a37c54b7575de695281c5ebb22
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434411
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: Jim Harris <james.r.harris@intel.com>
Add a check, which will be required for the further
unit test.
Change-Id: Ib1987fef914e6546f2bdbacd23bf9bb6005b8155
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Reviewed-on: https://review.gerrithub.io/435197
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
According to the TP 8000 spec, the maximal in capsule
data size is defined as follows:
1 For the Fabrics command and admin, it should not exceed
8192 bytes.
2 For I/O command, it shoudld be defined according to ioccsz
in the Identify controller data.
Change-Id: Ic13eda33e1516858e1e8749ee89459e3148d9e37
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/435826
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This is just a wrapper around the pmem_persist/pmem_sync
calls. It basically turns this:
if (vol->pm_file.pm_is_pmem) {
pmem_persist(buf, sizeof(buf));
} else {
pmem_msync(buf, sizeof(buf));
}
into this:
_reduce_persist(vol, buf, sizeof(buf));
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Id4e3f1538901cf7a3d5f5cec10b18907ca94afe0
Reviewed-on: https://review.gerrithub.io/434114
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This will be the logical block size presented by the
compressed volume to differ from the backing device's
block size.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie4ef06e131d8e101a0c9ced228c56a02fcbfb7af
Reviewed-on: https://review.gerrithub.io/434113
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This can be derived from chunk_size and backing_io_unit_size
in the params, but saving this value explicitly in the vol
structure is helpful so we don't always have to calculate
it.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ic389afcf60984ea431a6d1c7523005a368547447
Reviewed-on: https://review.gerrithub.io/434112
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Each request will need a scratch buffer of size
chunk_size. This is needed for read/modify/write
operations when only part of a chunk is written.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ided33f1e9ae18dd9a5de45f53f0a994a6f260b17
Reviewed-on: https://review.gerrithub.io/434111
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Right now a controller with a duplicated name will likely
fail to create any bdevs (as those can't have duplicated
names), but will still attach successfully. There will be
two controllers with the very same name and while this
doesn't seem to cause any data corruptions, it introduces
slightly non-intuitive behavior. After all, the controllers
are identified by their name and those should be unique.
This wasn't a major concern until we allowed creating
NVMe controllers without any namespaces.
Change-Id: I55dd67ef0b4e8a23f19269f9967109c4f54aec95
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434316
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
We will keep the RPCs for now but mark them deprecated.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I0407dcb392ea0c9e89c0f26cd5670aed2dbfadef
Reviewed-on: https://review.gerrithub.io/435345
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I90e7d698cae7577736319e38f089e3b759c9beef
Reviewed-on: https://review.gerrithub.io/435343
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
"trace" is for event tracing. SPDK used to use this
term for logging - we've moved some APIs to use "log"
but more needs to be changed. So start that now.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ib732c57d01602e56f37e9deed7135840a7c005be
Reviewed-on: https://review.gerrithub.io/435342
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
spdk_nvme_transport_id_parse() does not recognize the
namespace id, "ns", key as part of the transport id string
and thus logs an error message, but does not fail the call.
However, some SPDK applications, e.g. nvme/perf, in addition
to using spdk_nvme_transport_id_parse() also check for the
existence of a "ns" key in the transport id string to limit
the target to a specific namespace. This commit adds a
special case to spdk_nvme_transport_id_parse() to silently
ignore the presence of a "ns" key without logging it as an
error.
Change-Id: I49732b4d1b0227a38bb308eab1f6324dd241a2de
Signed-off-by: Lance Hartmann <lance.hartmann@oracle.com>
Reviewed-on: https://review.gerrithub.io/435192
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
I don't think this is a real problem but in stepping through
similar logic in the new compress vbdev I noticed that the
call to spdk_vbdev_register() which is now in the name search
loop due to a very recent code cleanup, will result in immediate
calls to this very same examine function and when unwinding will
continue through the loop for no good reason (a match was found).
Change-Id: I01583d10106008f1f75d5b3ecc7b64639e93d919
Signed-off-by: paul luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/435553
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This notice was scaring a lot of people because every time we disconnect
a qpair it tells the user that qpair is entering an error state. That is
part of the normal state flow of qpairs during disconnect, but makes it
seem like something is going wrong.
Change-Id: I776e71db2b24fa963113fee88b5cf02c0820f171
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/435555
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
DPDK already prints at least one error message, so
there's no need to print a yet another one.
Change-Id: I1c7bdfe5ca2095b93ec282bf193a717627d5fa27
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434410
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Prepare for storing additional per-device data.
The struct doesn't store any interesting data yet,
but already has a TAILQ_ENTRY that allows us to
put it into a global pci device list. Right now
we use the list only to find the SPDK device once
the corresponding DPDK device gets removed, but
more usages will be implemented soon.
Change-Id: If3abc1da60446e0a647d8d4c642f111ebfbcdb9e
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434409
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Now that even DPDK 16.11 (LTS) reaches its end of life in
November 2018, we can surely drop support for DPDK
versions older than that.
The PCI code will go through a major refactor soon, so this
patch cleans it up first.
Since this is the very first SPDK patch that drops support
for older DPDK versions, it also introduces an #error
directive that'll directly fail the build if the used DPDK
lib is too old.
Change-Id: I9bae30c98826c75cc91cda498e47e46979a08ed1
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/433865
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This case isn't particularly supported, but still
caused a memory leak and rendered the pci device
inaccessible for the rest of the primary process
lifetime.
This happens when a controller is removed from the
primary process while a secondary process still
uses it. The controller will likely misbehave without
its primary process managing it, but at least there
won't be a leak.
Change-Id: I67581cffa33ce14ff516b5743d13c9ef7b351625
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434408
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
With various possibilities to leak the rte_pci_device in the
primary process, we could technically construct the controller
in secondary. The nvme stack is not prepared for this and
will fail to initialize the device, but will still leak the
device object memory.
This patch adds an extra check to prevent any controller from
being constructed in secondary process.
Change-Id: I772f42b541c5db53310362b6595cebf9a30e8491
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434407
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Previously, if want to know which mask bit is used for specific
trace group, the only way is to check source code. Now list
each trace group with its trace tpoint group mask bit in
usage message
Change-Id: I7a85fe9c0885f1919f6ffbdc97dab81f1986fb07
Signed-off-by: Liu Xiaodong <xiaodong.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/435448
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
They used to be deleted together with the last NVMe bdev
built on top of them, but that was changed recently.
Currently controllers that aren't explicitly deleted are
leaked on lib finish.
While here, cleanup the destruct flag behavior and add
asserts against destroying the same controller twice.
Change-Id: I58878664602268398730fa4f619c2acd222317c9
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434317
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
We assumed spdk_mem_map_translate() translates only 2MB-aligned
addresses, but that's not true. Both vtophys and NVMf can use it
with any user-provided address and that breaks our contiguous memory
length calculations. Right now each buffer appeared to have the
first n * 2MB of memory always contiguous.
This is a bugfix for NVMf which does check the mapping length
internally. It will also become handy when adding the similar
functionality to spdk_vtophys().
Change-Id: I3bc8e0b2b8d203cb90320a79264effb7ea7037a7
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/433076
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
This is necessary to confirm that a buffer that spans a 2_MB boundary is
still in a single MR.
Change-Id: If0d14e514ab2197a0d2e3af4f565f56d50591210
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/435179
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
This keeps us from having to deal with ALLOC and FREE events
for mismatching regions - which necessitated splitting new
regions into individual pages. This caused all kinds of
problems with NVMe-oF - for example, buffers that spanned
memory regions, or bumping up against MR limits on RDMA
NICs.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I18dcdae148436b55d4481bb9fb8799f4832c7de1
Reviewed-on: https://review.gerrithub.io/434895
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
This patch disables the header digest and data digest
by default, which will improve the performance.
In the another patch, we will make it configurable.
Change-Id: Icdf8cda28217ec35a6b87bb932cdb1e4f8492471
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/435209
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This will be required in following histogram patches.
Signed-off-by: Piotr Pelplinski <piotr.pelplinski@intel.com>
Change-Id: I2eee6629243b7a4838a80dc1de33ae485c58081e
Reviewed-on: https://review.gerrithub.io/433874
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
NVMf target can't support extended LBA format for now, so print a
error log for those NVMe backend devices with extended LBA format.
Fix the issue #497.
Change-Id: Idda76ba934dd0eb45f92ae22b0b71398b3ae69dd
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/432799
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Minor cleanup just to make sure they are consistently <= 0.
Change-Id: I8427fd201e60e3f8ebbcf4929eb58ca164910623
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434324
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Currently spdk_bdev_part_base_get_tailq(gpt_base) will
return the global gpt tailq containing all the gpt part
bdevs, which is not what callers of this function expect.
Although the spdk_bdev_part_base_get_tailq() is currently
unused for gpt parts, it's still worth fixing it to make
the behavior consistent with other part bdev modules.
Fix this by having per-gpt-base tailqs which contain only
associated gpt partitions.
Change-Id: Ib3c4286fcc6912f2a252beb5b3dcafc0e5316434
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434836
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Currently spdk_bdev_part_base_get_tailq(split_base)
will return the global split tailq containing all
the split bdevs, which is not what callers of this
function expect. E.g. the construct_split_vbdev RPC
returns all split bdevs rather than the ones just
created.
Fix this by having per-split-base tailqs which
contain only associated splits.
Change-Id: I0fc25b28def0404f6a67152b5c21180e71660667
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434805
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Considering it's the part base object that's now accessible
in its remove callback, we can simplify the part API by making
it accept the part base object directly.
Change-Id: I87c3278929a063c115828d02e0def7fa536e6682
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434835
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Currently in the base bdev remove callback we don't
have access to anything but the spdk_bdev that's
being removed. Subsequent patches require the access
to more than that - e.g. some local metadata related
to that bdev.
By passing the part base object, we automatically get
access to e.g. spdk_bdev_part_base_get_ctx - a context
tightly associated with the part base, which can be
anything the upper layer (vbdev module) sets up.
Change-Id: Ifb99323978ef71ff6dd3b4ebf84fd21ef2920eb8
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/434834
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Application Makefiles can now just add xx_MODULES_LIB_LIST
to SPDK_LIB_LIST. This is possible now since all
SPDK libraries are linked with --whole-archive, so there
is no need to differentiate between "modules" libraries
and other SPDK libraries.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Iabf81a388b72d3b2a2f48287a8491ddc977722ac
Reviewed-on: https://review.gerrithub.io/434277
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
This allows a lot of simplification to SPDK application
makefiles.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I5fa463f6369834b84a8d92e79fa7768082209d7a
Reviewed-on: https://review.gerrithub.io/434274
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
There were several applications that were missing
either SOCK_MODULES_LINKER_ARGS or COPY_MODULES_LINKER_ARGS.
For the ones missing SOCK_MODULES_LINKER_ARGS (fuse and rocksdb),
the nvme inititator with TCP transport would not have worked
at all.
Adding COPY_MODULES_LINKER_ARGS to the bdev fio plugin enables
ioat which isn't critical, but adding it makes it consistent
with other apps and will allow its Makefile to be simplified in
some future patches.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I0098350d75d27ad2b2d408221b727698f5e902e4
Reviewed-on: https://review.gerrithub.io/434260
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
VHOST_USER_NVME_IO_CMD is designed to deliver NVMe IO command
header to slave target via socket, this can be used in BIOS
which will not enable Shadow Doorbell Buffer feature, since
we enabled the shadow BAR feature to support some old Guest
kernel without Shadow Doorbell Buffer feature, so the message
isn't required, just remove it.
Change-Id: I72e55f11176af2405c8cc09da404a9f4e5e71526
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/420821
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
For some old Linux Guest kernels, the new NVMe 1.3 feature: shadow
doorbell buffer is not enabled, while here, make a dummy BAR region
inside slave target, when Guest submits a new request, the doorbell
value will be write to the shared memory between Guest and vhost
target, so that the existing vhost target can support both new
Linux Guest kernel(newer than 4.12) and old Guest kernel.
Also, the shared BAR space can be used in future which we can move
ADMIN queue processing into SPDK vhost target, with this feature,
the QEMU driver will become very small and easy for upstreaming.
Change-Id: I9463e9f13421368f43bfe4076facddd119f4552e
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/419157
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
With the trace_size records for each lcore, spdk_trace
can read trace_file in which each lcore has different
number of trace entries.
Offset of each trace_history from the beginning of this
data structure.
Change-Id: I06afaba129812fe40ed000265fc66b02c5d9e3d9
Signed-off-by: Liu Xiaodong <xiaodong.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/433503
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Number of trace entries in circular buffer per lcore can be
assigned by the boot parameter of SPDK app with
"--num-trace-entries <NUM>"
Change-Id: I855ce6b4f14a716dcdd9078913da5ea8e577af3a
Signed-off-by: Liu Xiaodong <xiaodong.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/433099
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
vbdev registration was happing in a loop in 2 places immediately
following a call to claim(), moved the registration into the
function.
Change-Id: I880dccae02ac0262558119265d2940d0adca33dd
Signed-off-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/433727
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Initial implmentation had a 1:1 session to crypto op ratio. After
working with a similar concept in CompressDev checked to see if
this was required and apparently it is not.
Saves a decent number of API calls per crypto op and in the poller.
Also saves on mempool usage. Performance improvement measurement
is WIP.
Change-Id: I73f2355e720a16fd46bc4a02657419f779f07cbb
Signed-off-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/433726
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Fiona Trahe <fiona.trahe@intel.com>
Althrough it has very small chance to be executed, it's nice
to have it fixed.
Change-Id: I899681ccc13ed59c7fdd343ef7791df4e69e490f
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/433976
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
The ioctl NBD_SET_SOCK can return EBUSY if the kernel module
hasn't loaded entirely yet. Wait for it to become ready.
Change-Id: If3c6d0d8bb678ef8cab0efc1c5e800e95e19133e
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/433939
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This is going to need to wait for the kernel module
to load in the future, so split the function in
half.
Change-Id: I872d000acd4fc25737d5f50c8e0ae33641a6d7fa
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/433938
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Add a callback to spdk_nbd_start so that it can complete
asynchronously. As of this patch, it always calls the
callback immediately.
Change-Id: I6156fb203145362afa5e4102183b6cf143051c0c
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/433937
Tested-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Despite the scary commit title, this patch just unifies
per-driver mutexes into a single pci mutex.
On each hotplug we modify some DPDK global resources,
which per-driver locks aren't sufficient for. If
multiple threads try to attach devices at the same time,
then we'll likely have a data race. DPDK hotplug APIs
don't provide any kind of thread safety on their own.
Change-Id: I89cca9fea04ecf576ec5854c662bae1d3712b3fb
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/433864
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
We need to do it only for DPDK 16.11, which leaks the
mappings otherwise. DPDK was fixed in version 17.02 with
the following commit:
e84ad157 (pci: unmap resources if probe fails)
Unmapping the resources twice doesn't actually cause
us any trouble, but prints an ambiguous error message.
Change-Id: I8b62e86d5fff8fe924dbf9ae2e37cff29298d412
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/433863
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I5e7d16e377c03165f338709a71d6e4f03beffc0a
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-on: https://review.gerrithub.io/434066
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
According to the RFC3720, the length of initiator name and
target name should not larger than 223 bytes, while here,
add a check for initiator/target name.
Change-Id: I1517a4c4e1b0a944b239665f38f5dfb46f0075d2
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/433797
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
There are some use cases such as multipath and RAID expansion where a
vbdev could have been registered before one of its base bdevs.
Currently we unregister bdevs at shutdown in reverse order of their
registration. Continue to do that in general, but skip any bdev that
is still claimed. Any bdevs skipped in this way will eventually be
unregistered once any bdevs that have claimed it have completed
unregistration.
Change-Id: Iafde9558430bc5ce56e8608ef50bcb2b5fbfbf71
Signed-off-by: Andrey Kuzmin <akuzmin@jetstreamsoft.com>
Reviewed-on: https://review.gerrithub.io/432136
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Explicitly attaching a PCI device with spdk_pci_device_attach()
bypasses any kind of blacklists and should be only used
on a user request. Hotplug uevent is certainly not a user
request and should respect the blacklist, hence it's now
changed to call spdk_pci_enumerate() to probe new devices.
The enumeration callback will reject devices other that the
one we got hotplug request for, so no behavior is changed
in that matter.
This patch also fixes undefined behavior caused by reading
unitialized struct nvme_pcie_enum_cb;
Change-Id: I1399fbdd426152a13ed75c85a52bc7f0491ce287
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/433867
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
In addition UT were changed to set_thread() before registering
io device.
Change-Id: I959dbc800db9c1f50564274a73d71e05e843d8c9
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-on: https://review.gerrithub.io/433611
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
This matches the lib/ut_mock directory name, and also
avoids the final library being named spdk_spdk_mock.a.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I80ec09568d0455d1f131e45694b6e2d536e6ca91
Reviewed-on: https://review.gerrithub.io/433909
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
1) Add #define with comment for the 2-element iov array
2) Account for user passing trailing slash in pm file path
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Id5b620ae9d17dd3534e4cc78ddf1ba176f9c7fcd
Reviewed-on: https://review.gerrithub.io/433903
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
These structures will be used as contexts for every
I/O request. Allocate a bunch of these requests
when a volume is initialized or loaded.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ibf4da3081b03dd3950550e4ca1c5a3a50f72fd52
Reviewed-on: https://review.gerrithub.io/433516
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
This patch only creates and frees the bit arrays. Later
patches will initialize the bit arrays during load based
on information in the pm file.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Id3ba2b27fa7988778df440b3221ce12c80306c66
Reviewed-on: https://review.gerrithub.io/433515
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
These parameters will be needed for a number of reasons
after init/load.
The params struct contains the uuid, so the explicit uuid
member in spdk_reduce_vol is no longer needed.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ide12a7816e4a236be5c73a5733e76919ad9af103
Reviewed-on: https://review.gerrithub.io/433512
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
A pm file consists of:
* volume metadata
* logical map
* chunk maps
Define all three of these in struct spdk_reduce_vol.
Also define -1ULL to denote an empty entry in the logical
map or a chunk map - and initialize the logical map
and chunk maps entirely with this value when initializing
a new compressed volume.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9aae23c73d2fbbdc72050ab103fe9e686907eb40
Reviewed-on: https://review.gerrithub.io/433490
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
We will need the extra iov in the upcoming load path,
so we can load the superblock and pmem file path into
separate buffers.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I99cce0172bd772ca93813a002dcce6943e661f35
Reviewed-on: https://review.gerrithub.io/433087
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Currently the API requires the caller to open the
pmem file and pass the mapped buffer pointer, length
and pmem flag to libreduce using the spdk_reduce_pm_file
structure.
Let's have the library just do the pmem_map_file() instead.
Users then just pass the path to create the pmem file. The
library will still use the spdk_reduce_pm_file structure
internally - so move its definition out of the public header
and into reduce.c.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I81fcbfdfbb3211dca016d6aa422cf2e1ab16d84d
Reviewed-on: https://review.gerrithub.io/432593
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Previously, a pool of events per socket was created for NUMA
locality. However, when passing a message between threads,
the event would be taken from one pool and put back into another,
resulting in imbalances.
At this time, I do not see an efficient way to allocate the events
so that they remain NUMA local. We'll work on that over time.
However, for correctness right now, go down to a single global
pool of events.
Change-Id: I09f8b0c5c928777e8274c53c6dce21b9c346e2a5
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/433519
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
It is the first patch to follow the NVMe over fabrics
spec and implmenent the NVMe/TCP transport. It can be
divided into work in the host and target sides:
Host side: Add the TCP/IP transport in nvme lib (lib/nvme).
Target side: Add the TCP/IP transport in nvmf lib (lib/nvmf).
Change-Id: Idc4f93750df676354f6c2ea8ecdb234e3638fd44
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/425191
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This was dead code. Registering the same device more than
once has never been possible within a single process.
Change-Id: I04fa2a62cd9f3d7f99ff799ea4504c49a2232da4
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/433866
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Rework connect functions for upcoming non-blocking mode. As we are here
make some variables names shorter and more descriptive.
Change-Id: Ifcba24fc16f0931261067f6c2bf64eef450d0ec9
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/429912
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
As preparation for non blocking mode make the
spdk_jsonrpc_client_send_request functon just queue the request. Then in
spdk_jsonrpc_client_recv_response we actually send it. Without this
change send function will stay blocking.
Change-Id: Ida2290696d78afcb06d84a4538b74e2311043911
Signed-off-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Reviewed-on: https://review.gerrithub.io/429910
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Since the number of NVMe controller's Namespace is a const
variable, so the Namespace block device is a static array,
and is assigned with NSID after initialization, so when
users want to delete a specified block device, we can
just change it into inactive state.
Fix issue #495.
Change-Id: Ie6e33ad59f2375cc2b5099624bfe1ef6918a6264
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/433543
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Currently there are no timeout mechanism for Admin commands
when initialization, the NVMe driver may enter infinite loop.
While here, add a new parameter to the controller initialization
options, NVMe controller will report an error when timeout
happens during initialization.
Change-Id: Id0c6b6fa15abe5227b486bee95c8e02914b0d358
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/424622
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
The max_send_sge and max_recv_sge values can be set to any value from
0...dev_attr->max_sge. WHen we actually set the attributes, we will
receive a qpair with values for max_sge greater than or equal to what we
initially set. We need to store the maximum number of SGEs for later use
when constructing work requests.
Previously we have not relied on these values since we assumed that we
would always be able to have more sges than we asked for initially. This
may change as we try to allocate more SGEs to handle splitting buffers
across memory regions.
Change-Id: Ibbeae1908b86baa3a96d9c6cd2051401aaa2197b
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/433307
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
We have historically conflated SPDK_NVMF_MAX_SGL_ENTRIES and the maximum
number of SGEs associated with a wr data object. For now these are the
same thing, but there should be nothing tying the number of NVMe request
SGL elements to the number of rdma request wr sgl elements.
Also, clarify the rx_sge and tx_sge enums to reflect the actual maximum
number of SGEs associated with either the send and receive queues.
This change doesn't actually modify these values, but sets us up to do
things like split the data in an NVMe SGE into multiple WR SGEs in case
the buffer associated with the NVMe SGE is not contained in a single
RDMA mr.
We also need to store these values in the qpair for later usage.
Change-Id: Iff3756fc72787a4b72a99b2bdf90bf486a8010fa
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/433196
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
blob and blob_bdev are already included via the lvol bdev module
dependencies in mk/spdk.modules.mk. This patch adds blobfs there
too.
These changes are needed for some upcoming patches which will
apply --whole-archive to all libraries linked into the application.
That patch will require that we never include the same library
twice.
Shared library builds *require* that blobfs be listed before
blob on the link line, hence adding blobfs here even though it's
not technically associated with a bdev module.
We can also simplify spdk.modules.mk while we're here - just
filter out bdev_lvol to disable lvol for applications like
blob_cli that need to disable it.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I01e5e6b91d947463749f942f54c12f48229bbe5a
Reviewed-on: https://review.gerrithub.io/432916
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Same sequence used in recent crypto module fixes.
Change-Id: Ic5f714e91cf871bce35b0e8fe3e32ae42157e6de
Signed-off-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/433383
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Move the message ring and the polling into the thread
itself.
Change-Id: I5bc53c68bb1314fd69ed80cc98c5b86ce8be880f
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/423768
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Picked up from recent update to the passthru module.
Change-Id: Ie7404e5aadda6691b15020adbbc1e1f4c23d0a12
Signed-off-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/433384
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
All io_devices now have an associated name - so add
that name to relevant error messages to help with
debugging.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I8ef678681fc29daed08e2f95b3096c5dce429ab9
Reviewed-on: https://review.gerrithub.io/433203
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Having the buffers be the same size as the maximum xfer size doesn't do
us any favors. Make these buffers a ratio of the maximum transfer size
and the number of supported nvmf SGLs.
Also configure the number of nvmf request iovs to correspond with this
new ratio.
Change-Id: I3147dcd86b599c74521ebfdf3bcdbcdee8871a3a
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/428747
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
According to the currrent logic, it should be explictly defined
in the configuration file if the NVMe-oF target is started by
configuration file without using rpc calls. If no transport is
defined, we do not need to parse the following subsystems
but should terminate the NVMe-oF target .
Change-Id: I2e2db8406a30a9bf7e54d38f8d08a8d92ef158c9
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/432376
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
In vbdev_crypto_examine() we were failing to unregsiter the io_device
in the event that spdk_vbdev_register() call failed. Found via
inspection.
Change-Id: I73c6c0c5693777b93c1ea02dcf2e2e65d46fe27d
Signed-off-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/432933
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
In create_crypto_disk() we were failing to unregsiter the io_device
in the event that spdk_vbdev_register() call failed. Found via
inspection looking into a CI failure however this potentially could
have caused that failure as well (I don't think so though, there
were no prints in the log that it followed this path).
Change-Id: I7085c4e25665a5a15def38d6726d519731b7e44e
Signed-off-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/432932
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Found via inspection while invetigating a CI failure. In
vbdev_crypto_examine() we were not looking at the rc from
vbdev_crypto_claim()
Change-Id: I8be09b5844e18e35b95f19e378fe280323d183fa
Signed-off-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/432930
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The io device was previously not being unregistered, this looks
like the root cause to several recent CI failures in the bdev
JSON tests that use RPC to create/save/clear/load configs.
Change-Id: Ia77ed9fe230c79188d8d862e98b17581ae81b31f
Signed-off-by: paul luse <paul.e.luse@intel.com>
Reviewed-on: https://review.gerrithub.io/433194
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
All io_devices should be unregistered before the
application exits. If not, print an error message
for each io_device that still exists.
Eventually we may want to assert() if any io_devices
are found, but for now we'll just stick with the
error message.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: If017edf40fc2904ed772bc1c6124f06fc9b49bcd
Reviewed-on: https://review.gerrithub.io/433202
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Eventually we may want to assert if any channels still
exist when the thread is freed, but for now we'll just
stick with the error message.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I909a409f34c7783a32d3300942897f591c90fca1
Reviewed-on: https://review.gerrithub.io/433201
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
The user of the thread library is now only responsible for
periodically calling spdk_thread_poll. Pollers are handled
internally.
In order to avoid changing all of the unit tests, the ability
to provide function pointers to change the behavior of
the poller registration is still in the code. This should
only be used from tests until they are all converted.
Change-Id: Ie2c00ce1d57ca3710ed2c469cd711924768e23ef
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/417784
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Previously we used VFIO if only the vfio-pci kernel module
was loaded, which is different from what our setup.sh script
did. On a fairly usual system configuration, setup.sh could
have bound devices to UIO, but SPDK would still try to map
memory to an (empty) DPDK VFIO container just because its fd
was available. That would fail obviously.
setup.sh checks for IOMMU presence in order to use vfio-pci
and SPDK should probably do the same. We could check the
kernel driver of each attached PCI device, but there's no
chance right now of supporting both UIO and VFIO devices
at the same time with IOMMU passthrough. That's not
a reasonable configuration anyway. To keep things simple,
we just add a single check on vtophys initialization.
Fixes#462
Change-Id: Ica653f117743be322291a1b7e37ed00e34ef5035
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/432518
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: <wuzhouhui@kingsoft.com>
Write pm filepath first to offset 4K. Then write
the super block to offset 0. This ensures the backing
device isn't really valid until both are written
(avoiding the case where the super block got written
but not the path.)
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I55508aa67b4179e658827c982cd955d009a3f321
Reviewed-on: https://review.gerrithub.io/432505
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: <spdk-ci.pdl@broadcom.com>
While here, add basic implementations for these function ptrs
in the unit tests.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I54e18a3b331777602fed29382b95b449005efcce
Reviewed-on: https://review.gerrithub.io/432503
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Signed-off-by: Piotr Pelplinski <piotr.pelplinski@intel.com>
Change-Id: I621a8a3891c8f3829564940219c70c6520c6f9c7
Reviewed-on: https://review.gerrithub.io/427899
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Printing messages every time an iSCSI connection logs in or
logs out/terminates is too noisy. Make these DEBUGLOGs
instead.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ifc1014e4f8dce414a1c4ef75ea6b6749954083cc
Reviewed-on: https://review.gerrithub.io/432606
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
The bdev layer will sort of check this when we try to
register the bdev, but it's better to catch this much
earlier. That way we catch it when the name gets inserted
into the list, rather than when the base bdev appears.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I846cb138fa848078d2c76356273870c607d79548
Reviewed-on: https://review.gerrithub.io/432602
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This ensures any errors get properly propagated back to the
RPC handler.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia352b3bd7d8912e7095e9d3e6b2e80e3dccbeb42
Reviewed-on: https://review.gerrithub.io/432601
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
The RDMA memory map needs to be per-protection
domain, not per NVMe controller. Otherwise, when
an NVMe controller is removed, the memory map may
reference an invalid pointer to a detached
controller.
Change-Id: I0c5bd2172daee0c70efb40eab784839e0cde8bc4
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/432590
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Apparently this patch revealed an issue somewhere in SGL
handling in NVMf initiator which is causing our CI to fail.
Let's revert it while we work on a proper fix.
This reverts commit 6f5840d225.
This is not a full revert, some app.c changes to the usage text
were kept unchanged.
Change-Id: Iddea5c2b9df50bd12ef8f6226165883f6622ab33
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/432576
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: Ben Walker <benjamin.walker@intel.com>
In some special cases, NVMe device with cdata.nn=0
may be used to do validation or other test work.
cdata.nn=0 means the device can't support NS at all.
Change-Id: I55f75a8cb21b8d1b99c5318e27c876a4371d6dd4
Signed-off-by: Liu Xiaodong <xiaodong.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/432191
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: joevannip <jparairo@nvxltech.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
BSD implementation for config access in DPDK seems to
return 0 on success while Linux implementation returns 0
only on failure. The env wrapper was always treating 0 as
an error and caused some of our PCI initialization code
to fail prematurely.
At one point DPDK harmonized this BSD behavior with Linux,
but only for config reads.
Fixes#484
Change-Id: I4ea850ea50f5e667fad28e8125209b21c377a2a3
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/432401
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Now, that _spdk_bdev_io_get_buf offers allocating aligned buffers,
add possibility to store original buffer and replace it by aligned
one for the time of IO.
Signed-off-by: Piotr Pelplinski <piotr.pelplinski@intel.com>
Change-Id: If0ed306175631613c0f9310dccaae6615364fb49
Reviewed-on: https://review.gerrithub.io/429754
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Allow specifying a custom hugetlbfs directory.
This can be useful e.g. when trying to use hugepages
with fixed size, different size limit, or different
access permissions.
Change-Id: I418cbab99ed183383300b3c3d9945095a03478db
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/432105
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This option is deprecated. Also, rename the rpc and configuration
options for setting the opts to reflect that they now only set the max
number of subsystems
Change-Id: Iaabcbf33dd0a0dc489d81233fda74e9e7f3e0d2e
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/430161
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
In the usage output, the default config could be printed
as "(null)" and mem_size as "-1MB".
Change-Id: Ib6adc1ab86c43a40096d392bdfc64d3d508a86bc
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/432422
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Now that we utilize DPDK dynamic memory allocation, we
no longer need to set the default pre-reserved memory size.
It'll now be 0, which implies all memory will be allocated
at runtime.
The option to pre-reserve all hugepages on the system is
now only available on BSD, so also clarify that bit in the
app usage text.
Change-Id: I5a8a1d9bf14ad6d938532d7e6254a45e4a81bb92
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/432204
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This patch implements the following QP shutdown flow:
1. Move the QP to ERR state
2. Post dummy work requests to send and receive queues
3. Poll CQ until it returns dummy work requests (with WR Flush Error status)
4. Call ibv_destroy_qp and release resources
In order to differentiate dummy and normal WRs new spdk_nvmf_rdma_wr
structure was introduced which contains type of WR. Since now it is
expected that wr_id field in ibv_recv/send_wr and ibv_wc always points
to this structure. Based on WR type wr_id can be safely casted to
correct container structure. In case of unsuccessful work completions
'opcode' can not be used for this purpose because it may be
invalid (see "IB Architecture Specification Volume 1", ch. 11.4.2.1
"Poll for completion").
Change-Id: Ifb791e36114c619c71ad4d831f2c7972fe7cf13d
Signed-off-by: Evgeniy Kochetov <evgeniik@mellanox.com>
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Reviewed-on: https://review.gerrithub.io/430754
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Add the following to struct spdk_reduce_pm_file:
* path of the pm file
* pointer to the mmapped pm region
* whether the mmaped address is pmem or not
Now use pmem_persist or pmem_msync to persist volume
parameters. Note that we do *not* persist the pm filename
to the pm file - the pm filename will be written to the
backing disk in a later patch.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I1d873ae086dc20600798a49b98e1bf490ac750f3
Reviewed-on: https://review.gerrithub.io/430648
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
--with-reduce required to build reduce. This depends on
libpmem being installed.
We still need to work out details in pkgdep.sh and
vm_setup.sh. Some distributions like Ubuntu still
require configuring extra package repositories to
get libpmem packages.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I4e056ce1da9a1fecb4458f8f5e7ff5d61c422533
Reviewed-on: https://review.gerrithub.io/430646
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
A compressed volume will require both a persistent
memory region/file for per-chunk metadata and a
backing device to store the compressed blocks.
Add functions here to calculate the sizes of these
based on the desired size of the compressed volume,
its chunk size and the size of each backing block.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9203479e2a268c3ab0e2b0e06e348285e9d1cd13
Reviewed-on: https://review.gerrithub.io/430387
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
reduce will be a block compression algorithm designed
specifically for SPDK. It is called "reduce" because
it reduces the data size on disk.
This patch just adds the shell of a library, include
files and unit tests. This will be fleshed out in
the rest of the patches in this series.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I88e238af64142a7c0e50ab7b447280026b55581f
Reviewed-on: https://review.gerrithub.io/430386
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
NVMe controller won't be removed if it has no namespace
or no bdev is succesfully created out.
Change-Id: I45b7e364a0cb35c902af5e745237931b67335145
Signed-off-by: Liu Xiaodong <xiaodong.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/431701
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
