This will allow us to move more of the reset-related
functionality to the common layer, as part of enabling
resets for fabrics controllers.
The transport qpair_enable and qpair_fail functions
acted similarly - so those are both removed now and
replaced with this new qpair_abort_reqs function.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9486630ad5b807239b0b5bcde50e8cfd313695d3
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453486
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: Changpeng Liu <changpeng.liu@intel.com>
Upcoming patches will move the actual is_enabled
logic to the common layer as well.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9fd4d8712280295f57134ad66f8ccbfe9736d30a
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453484
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: Changpeng Liu <changpeng.liu@intel.com>
We submit AERs to all controllers - both pcie and
fabrics. But currently we only manually abort the
aers when disabling the qpair for pcie. Make this
common instead by creating a new transport function
for aborting aers.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I1e926b61b8035488cdc6e8cb4336b373732f985e
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453482
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: Changpeng Liu <changpeng.liu@intel.com>
This better explains what the function is doing,
and makes the name more general so we can use it
for the adminq as well.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I6b55761cb141a9a79cdef876be47995d8813b312
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/453480
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Send the given NVM I/O command, I/O buffers, lists and all to
the NVMe controller.
This is a low level interface for submitting I/O commands directly.
This can only be used on PCIe controllers and qpairs.
This function allows a caller to submit an I/O request that is
COMPLETELY pre-defined, right down to the "physical" memory buffers.
It is intended for testing hardware, specifying exact buffer location,
alignment, and offset. It also allows for specific choice of PRP
and SGLs.
The driver sets the CID. EVERYTHING else is assumed set by the caller.
Needless to say, this is potentially extremely dangerous for both the host
(accidental/malicionus storage usage/corruption), and the device.
Thus its intent is for very specific hardware testing and environment
reproduction.
Signed-off-by: James Bergsten <jamesx.bergsten@intel.com>
Change-Id: I595fe02fe0dfa9c3ceba1ac116b6900357b02d2c
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/451994
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 mode of dmb oshld can guarantees cpu sequential execution,
which has less impact on performance.
Change-Id: If30b6a682a2216eecd1da039267ed4f5471afc38
Signed-off-by: h00448672 <heyang18@huawei.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/446827
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 an attempt to fix device hotremove with VFIO.
A soft device hotremove request through sysfs [1] would
currently just block until the SPDK process manually
releases that device - e.g. upon an RPC request.
VFIO won't get unbound from the device untill userspace
releases all its resources. VFIO can signal a pending
hotremove request by kicking any file descriptor provided
by the userspace - and DPDK does provide such descriptor -
but SPDK does not listen on it.
DPDK does offer handy API to listen and in this patch
we make use of it inside our env/pci layer. Within
a DPDK callback we set an internal per-device hotremove
flag, which upper-layer SPDK drivers can poll with a new
env API - spdk_pci_device_is_removed().
The VFIO hotremove event will be sent to primary
processes only, so that's where we listen.
We make use of this new API in the NVMe hotplug poller,
which will process it just like any other supported
hotremove event.
Fixes#595Fixes#690
[1] # echo 1 > /sys/bus/pci/devices/<bdf>/remove
Change-Id: I03d88271c2089c740e232056d9340e5a640d442c
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448927
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This avoids a data dependent load to find which
callback to call in the completion path.
Change-Id: Ifa20790a7af3332a74bc45037e589668744af797
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/450558
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>
The same information can be obtained by checking the req
field for NULL.
Change-Id: I9689ee0be33537fd6d3f35d8c3710fabab5e1928
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/450557
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: Changpeng Liu <changpeng.liu@intel.com>
When copying the NVMe command from the request to the actual
submission queue slot, use a non-temporal move instruction.
The submission queue slots are never read by software - only
written to. So don't pollute the CPU cache with their contents.
Change-Id: I112f721abfac03bd7b33ec9ddf783d4bf2952b42
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/450193
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Xiaodong Liu <xiaodong.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
There are several places where we have the tracker
pointer, yet we go find the tracker again by getting
the tr->cid and using that index to find the tracker
again in the qpair's array. That's really silly.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I54acd642a2c9821f2b95e17563904b859495081a
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/450308
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Logically we should complete any I/O in the completion queue and
abort the rest of the I/O after we delete the submission queue
and completion queue, so that we would not lost any complete I/O.
We alse should complete I/O and abort I/O before destroy I/O qpair
even though the ctrlr/device has been removed.
Change-Id: Ieb28ad7b4a3a7be553f70178b29ca870b5413191
Signed-off-by: JinYu <jin.yu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/449316
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>
Similar to recently added nvme_pcie_qpair_ring_sq_doorbell.
Prepares for using this code for coalescing cq doorbell
writes in the future.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I565e103acf73b3d305e72a4440e1cc678c95faa0
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448871
Tested-by: SPDK CI Jenkins <sys_sgci@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>
The phys_addr param in spdk_*malloc() is about to be
deprecated, so use a separate spdk_vtophys() call to
retrieve physical addresses.
This patch also adds error checks against SPDK_VTOPHYS_ERROR.
The error handling paths are already there to account for
spdk_*malloc() failures themselves, so reuse them in case
of vtophys failures.
Change-Id: I377636e66b8c570d013c1bb2021f04bce4e6c0ce
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/416998
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>
Try to group data members that are used often into the
same cache lines. We still need to find more space in the second
cache line of spdk_nvme_pcie_qpair so that the important
parts of spdk_nvme_qpair fit.
Change-Id: Ib936cb2b1acc722de7ec313d6faa3812aacde394
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447968
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>
Don't touch the shadow doorbells if it isn't necessary.
The flag could be combined into a bit mask with other
flags in a future patch.
Change-Id: I9ffd16468d29f0f0868cf849f7fece327eb6a294
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447967
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>
Avoid ringing the submission queue doorbell until the
call to spdk_nvme_qpair_process_completions().
Change-Id: I7b3cd952e5ec79109eaa1c3a50f6537d7aaea51a
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/447239
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>
This is going to get called from two places shortly.
Change-Id: I2c67e719c91887987e6e65c5c0c384bed0431409
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/448311
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Move req->submit_tick assignments from specific transports to generic
qpair code.
Check whether submit_tick has been assigned before doing the actual
assignment, because a request may be submitted several times and the
original submit_tick shouldn't be covered.
Change-Id: I2de8018dc21763eb5a19bb9d48dfbdef764b036e
Signed-off-by: lorneli <lorneli@163.com>
Reviewed-on: https://review.gerrithub.io/c/444702
Reviewed-by: Ziye Yang <ziye.yang@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>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Add a memory barrier for arm64 to prevent possible reordering
of tracker and cpl access,
because arm64 has less strict memory ordering behavior than x86.
Change-Id: I0a8716f7bfeffb0bbce27ee3174e214c8e4566b4
Signed-off-by: heyang <heyang18@huawei.com>
Reviewed-on: https://review.gerrithub.io/c/442964
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.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>
Existing NVMe driver uses a global list g_nvme_init_ctrlrs
to track the controllers during initialization, and internal
function will start each controller in the list one by one
until the list is empty. We introduce a probe context
and move the global list into the context, with the context
we can enable asynchronous probe API in the next patch, also
this can enable parallel probe feature.
Change-Id: I538537abe8c1a4a82fb168ca8055de42caa6e4f9
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/c/426304
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>
Currently infrequent cases in request completion path are marked as
unlikely. This patch applies that to submission path.
These cases are infrequent and marked using unlikely marco:
a. The sq tail reaches the end of queue.
b. The sq tail equals to sq head. (never happen if FW runs correctly)
c. The qpair is admin queue.
Change-Id: I8b873a18615788f2efbf7c683aad710c7007a082
Signed-off-by: lorneli <lorneli@163.com>
Reviewed-on: https://review.gerrithub.io/c/443451
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Makes the code slightly more readable.
Change-Id: Iebf8fb07bceacf433d4bdad0a30419a3faab7eee
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/439370
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 use those values in various places in SPDK,
so let's define them in a single place now.
Change-Id: Iad9a5745d69166a6e6032370d4e5a0e604914e45
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/439369
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>
This follows the same trend as the mem_map APIs.
Currently, most of the spdk_vtophys() callers manually
detect physically noncontiguous buffers to split them
into multiple physically contiguous chunks. This patch
is a first step towards encapsulating most of that logic
in a single place - in spdk_vtophys() itself.
This patch doesn't change any functionality on its own,
it only extends the API.
Change-Id: I16faa9dea270c370f2a814cd399f59055b5ccc3d
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/c/438449
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: wuzhouhui <wuzhouhui@kingsoft.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
As a part of cleanup they're replaced by a device-agnostic
attach API, which is easier for us to manage.
Change-Id: I7558590e41e5c580a130a6aba7ae4f7dcff58da8
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/436478
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>
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>
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>
The spdk_dma_zmalloc guarantee about physical memory contiguity
is about to be removed soon. For hardware rings that require
physical memory or IOVA contiguity we will now enforce hugepage
alignment and size restrictions to make sure they occupy only
a single hugepage.
Change-Id: Iebaf1e7b701d676be1f04a9189201c5d89dad395
Signed-off-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/418547
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Chandler-Test-Pool: SPDK Automated Test System <sys_sgsw@intel.com>
A few open-coded sequences equivalent to SPDK_CONTAINEROF() were
scattered around; replace them with the macro from spdk/util.h.
Change-Id: I95c6e6838902f411420573399ced7c58c2e4ef84
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/418126
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
The spdk_dma_zmalloc guarantee about physical memory contiguity
is about to be removed soon. A single tracker is page size
aligned and is exactly one page big, so it is physically
contiguous, but we can't assume an array of those is physically
contiguous as well.
Change-Id: I3aa4d14dd677601c30aa2d8f15197886d6c46e58
Signed-off-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/416840
Tested-by: 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: Daniel Verkamp <daniel.verkamp@intel.com>
There's no need to split a bufffer if it's physically
contiguous. We can now merge buffers that would be
previously split by the nvme_pcie driver and also
separate SGEs provided by the user that happen to be
physically contiguous.
Change-Id: I9c9de31d52a9dc9e384806555cb94609aff0ccf3
Signed-off-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/417061
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
This allows NVMe PCIe devices to be used with
physically discontiguous I/O payload buffers.
So far this is just a dumb splitting which
doesn't check for physical contiguity. This is
improved in a subsequent patch.
Change-Id: I0ecc443149225eaa0e4156ddda78613bcf034406
Suggested-by: Daniel Verkamp <daniel.verkamp@intel.com>
Signed-off-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Reviewed-on: https://review.gerrithub.io/417060
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Fix issue #313.
For multi-process test scenarios, the secondary process may access
controller’s CSTS register in the shared ctrlr list. For this situation,
all the controllers are already in the primary shared ctrlr list, but
then each controller is added one by one in the secondary process, so the
secondary process may access CSTS before it is remapped for the BAR space.
In the rpc_config.sh test case, the spdk_nvme_ctrlr_get_regs_csts function
will be called in _nvme_pcie_hotplug_monitor function before calling
spdk_pci_nvme_device_attach. This step caused the secondary process iSCSI
Target access CSTS before it is remapped for the BAR space.
Change-Id: Ifd62c38adf8624f9877a9a2f965ca4db28839d99
Signed-off-by: Liang Yan <liang.z.yan@intel.com>
Reviewed-on: https://review.gerrithub.io/412594
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: GangCao <gang.cao@intel.com>
Reviewed-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Factor qpair destruction function so that we can put common
resource release together in future.
Change-Id: I44139947820c2a384b745ae2673799f1b736369c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/412604
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This will be used in other transports as well.
Change-Id: I05026b0dfea2647d61a173379aca368ca48a2f52
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/413864
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
If requests are in flight at the time a timeout callback is configured,
we can't retroactively get the submission time of those requests; treat
them as not having a timeout.
Change-Id: Ic589e874a0f2c8c7f4ce352afa7c3aea33b01fae
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/413863
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
This is the first step toward timeout handling for other transports.
Change-Id: I386dd990f667d449e94ba4bcedaa3435743755fd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/413862
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
This was partially fixed in commit ddeaeeec19 ("nvme: Only check
timeouts on requests from the same process"), but the function that
calls nvme_pcie_qpair_check_timeout() was also erroneously filtering out
the admin queue. Restore the original behavior of checking all queue
types.
Change-Id: I26a44ff5eb772735d314ce7b8322ba9222675911
Fixes: 31bf5d795e ("nvme: make timeout function per process")
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/411628
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
When the shadow doorbell feature is enabled, no actual MMIO takes place
when a shadow update is sufficient; tighten the bounds of the updates to
g_thread_mmio_ctrlr in the two doorbell update locations so that we only
need to touch the thread-local variable when actual MMIO access is
required.
Change-Id: Ida974bec33f56cbb9f7d3611f483c6975ec773ab
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/413856
Tested-by: 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: Jim Harris <james.r.harris@intel.com>
Rather than storing nvme_payload::type explicitly, use the SGL reset
function pointer as an indicator: if reset_sgl_fn is non-NULL, then the
payload is an SGL type; otherwise it is a contiguous buffer type.
This eliminates the one-byte type member from struct nvme_payload,
making it an even 32 bytes instead of 33, allowing the removal of the
awkward packing inside struct nvme_request.
Change-Id: If2a32437a23fe14eb5287e096ac060067296f1dd
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/413175
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This will simplify upcoming patches that change the way nvme_payload
stores its type.
Change-Id: Idf0a5b8dfd7d66a10f89254d2c5c54fee2968a43
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/413173
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
For some cases, especially for Admin commands, there maybe has
recursive commands, e.g.: in AER callback we may send a new AER
request, in such case, the current code can't process such
case. While here, move the completion queue head to next before
any response callback will fix such issue.
Change-Id: Ide56701d94615881790cf025ede2f07420b9b16e
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/412766
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Factor out the common pattern of waiting for an internally-submitted
command to complete. This will give us a convenient central place to
add error checking.
Change-Id: I65334d654d294cfb208fc86d16fa387ac5432254
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/412545
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
When IO is finished SPDK will trigger callback at controller layer,
while here, wrapper the completion callback into a function so
that we can add error injection at this function in following patch.
Change-Id: I7b7a6d278d87fd09a05f51f688398fdf2e9c4e05
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/411630
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I7598222db5d76c1a1578fbb5935d4348f7c62f54
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/410951
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This will be used later instead of retrieving VS (potentially via a Get
property command for Fabrics) multiple times.
The Active NS List code was previously depending on the VER field of the
Identify Controller data, but this was only added with NVMe 1.2, so we
can't rely on it to detect NVMe 1.1 controllers; it is changed to use
the new cache VS value instead.
Change-Id: Iba9ed5ecbc82b4654973438d119daba0c4cf0724
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/408895
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Change-Id: I80521c4c01daf033319f88cf273255387a7b5248
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/408403
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This fixes the VFIO hot-remove path, which called remove_cb without
checking to see if it had been specified by the user. The normal uevent
removal path already checked for remove_cb.
Change-Id: I0ad8d2c90a77b16800a8b505cb69ea05b0706d70
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/408392
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This was broken by commit 31bf5d795
Change-Id: I8c81c7b76cd47db347ce9c3f8a0e8296b690cb49
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/408240
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9fee91ece79b204962a70fc49d9032abe2c55090
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/408218
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I6e58baaeb09580b5f70e1acf5323376ca0b26bbf
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/407382
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Automatically detect more whitespace errors.
All existing cases are fixed; only whitespace change (verify with
diff -w) except for one comment style fixup in include/spdk/nvme.h.
Change-Id: If750e54b9c8e3421ea6feda5f20184a31431631e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/402360
Tested-by: 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>
There was a bug reported by Cunyin Chang with regards to how the
cmb_current_offset was calculated when the CMB offset into the CMB BAR
is non-zero. This patch fixes this issue and also fixes the problem
that the last valid offset into the BAR *may* differ if registration
is utilized or not (due to the 2MiB alignment and length requirements
for registered memory).
Change-Id: Id08d6a5a40b828338f6a66599171cc8dd59768a3
Signed-off-by: Stephen Bates <sbates@raithlin.com>
Reviewed-on: https://review.gerrithub.io/401832
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
adding nvme_ctrlr_destruct_finish because nvme_transport_ctrlr_destruct may
use a destroyed mutex.
nvme_ctrlr_destruct() free "ctrlr_lock" and after that call
nvme_transport_ctrlr_destruct()->nvme_pcie_ctrlr_destruct()(with pci)->
nvme_ctrlr_proc_get_devhandle()->nvme_robust_mutex_lock(&ctrlr->ctrlr_lock);
Change-Id: I55714ea9097d2c9d844a00b5a88fa2d51a3f4469
Signed-off-by: Ehud Naim <ehudn@marvell.com>
Reviewed-on: https://review.gerrithub.io/399605
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I2a3c7a272dc08be5a5ecb4339622816482c4cbb0
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Signed-off-by: Stephen Bates <sbates@raithlin.com>
Reviewed-on: https://review.gerrithub.io/397036
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Enable address translation for I/O buffers within the controller memory
buffer region by registering the CMB using spdk_mem_register().
Change-Id: I44829757ad15fbc3ea96fa494b9fb32dd67a7138
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Signed-off-by: Stephen Bates <sbates@raithlin.com>
Reviewed-on: https://review.gerrithub.io/397035
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
For a given hardware queue size, only allow a quarter of the queue size
to be returned as completions in a single call to
spdk_nvme_qpair_process_completions(), and adjust num_trackers to match
so that num_trackers + max_completions_cap doesn't exceed the hardware
completion queue size. This ensures that there is room in the
completion queue if new I/O is issued in response to completions before
we ring the completion queue doorbell.
The choice of 1/4 queue size is arbitrary; this seems to be a good
compromise between completion batch size and number of trackers.
Change-Id: I2c5aad7b98bfc8b33e53242240b2c9254fa05b4e
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/393529
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Commit ID "269910c0" removed the support of separate metadata,
for those controllers which can support this feature, SPDK driver
can't be used. SPDK provides APIs such as:
spdk_nvme_ctrlr_cmd_io_raw_with_md/spdk_nvme_ns_cmd_write_with_md/
spdk_nvme_ns_cmd_read_with_md, which can support separate metadata.
While here, re-enable this feature with this commit.
Change-Id: If77c21e9ac700c4b334548ebfa7e8e6286285a64
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/392440
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I8519a4b68db44cb8fe6dd251a52bf0f1dca73c32
Reviewed-on: https://review.gerrithub.io/391890
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Require braces around all conditional statements, e.g.:
if (cond)
statement();
becomes:
if (cond) {
statement();
}
This is the style used through most of the SPDK code, but several
exceptions crept in over time. Add the astyle option to make sure we
are consistent.
Change-Id: I5a71980147fe8dfb471ff42e8bc06db2124a1a7f
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/390914
Reviewed-by: <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Tested-by: 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: Jim Harris <james.r.harris@intel.com>
Disambiguate the log components from the trace functionality
(include/spdk/trace.h).
The internal spdk_trace_flag structure and related functions will be
renamed in a later commit - this is just a find and replace on
SPDK_TRACE_* and SPDK_LOG_REGISTER_TRACE_FLAG().
Change-Id: I617bd5a9fbe35ffb44ae6020b292658c094a0ad6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/376421
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
NVMe specification 1.3 added a new Admin command: Doorbell buffer config,
which is used to enhance the performance of host software running in
Virtual Machine, and the Doorbell buffer config feature is only used
for emulated NVMe controllers. There are two buffers: "shadow doorbell"
and "eventidx", host software running in VM will update appropriate
entry in the Shadow doorbell buffer instead of controller's doorbell
registers.
Change-Id: I639ddb5b9a0ca0305bf84035ca2a5e215be06b46
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/383042
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This patch just make sure SPDK support the hotplug of vfio-attached devices,
but there still issues when we get hotplug events from kernel, this patch tested
based on my modified local vfio-pci driver which just for test purpose and not upstream.
Change-Id: I974177e70e4133d2a583d00e714a7607765d3548
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
Reviewed-on: https://review.gerrithub.io/390580
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Simplify the PCIe transport by using an existing function to look up a
controller by transport ID.
Change-Id: I261865df1ba23069b052ca64944b7637d70c85ba
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/388701
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Because ppc64 has less strict memory ordering behaviour
than x86 we need to introduce a memory barrier when
polling for completions due to possible reordering
of tracker and cpl access.
Change-Id: Id17116c38b2ba69154c175c539fc97c60897deb0
Signed-off-by: Jonas Pfefferle <jpf@zurich.ibm.com>
Reviewed-on: https://review.gerrithub.io/383728
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
cmd and cpl rings must be aligned on page size boundaries.
Change-Id: I103eed32adfaa4388bc7d672ee166973f796b343
Signed-off-by: Jonas Pfefferle <jpf@zurich.ibm.com>
Reviewed-on: https://review.gerrithub.io/383727
Reviewed-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
spdk_pci_device_claim() can be used to ensure only one process
at a time uses any given PCI device. Previously this was only
used in the bdev_nvme driver - other apps like nvme/perf do
not use spdk_pci_device_claim() and could effectively rip out
the device from a running bdev-based app like the NVMe-oF target.
So instead of modifying all of the nvme apps, put this logic into
the core nvme driver instead so that all applications get the
benefit transparently. Save the fd when the controller is constructed
and then close it when the controller is destructed to handle the
detach (including hotplug) cases.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I5dc48a2e41dc06707800f15a9e1f9141477628c6
Reviewed-on: https://review.gerrithub.io/385524
Reviewed-by: Dariusz Stojaczyk <dariuszx.stojaczyk@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
With this new API, callers can attach one specific ctrlr identified by
the transport ID directly along with optional ctrlr opts. If connecting
to multiple controllers, it is still suggested to use spdk_nvme_probe()
and filter the requested controllers with the probe callback.
Two primary use cases:
1) connecting to the NVMe-oF discovery controller
2) more straightforward way to connect a specific controller (avoiding
the probe callback)
A typical usage of this API with specific ctrlr_opts:
1. struct spdk_nvme_ctrlr_opts user_opts = {}
2. Call spdk_nvme_ctrlr_get_default_ctrlr_opts(&user_opts, sizeof(user_opts))
3. Modify the content of the initialized user_opts with user required value like
user_opts.num_io_queues = 8
4. Call spdk_nvme_connect(&trid, &user_opts, sizeof(user_opts))
Change-Id: Idf67ee5966f6753918c12604342c892d2f3bbe3a
Signed-off-by: GangCao <gang.cao@intel.com>
Reviewed-on: https://review.gerrithub.io/370634
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
SPDK already uses DEFAULT_IO_QUEUE_SIZE and MQES to decide the correct
queue depth of NVMe queue pair, hardcoded it to NVME_IO_ENTRIES(512)
does not make sense if users want to set queue depth bigger than 512.
Change-Id: Iaa73fc79e055292ae9bd19af0c8c12f257ae5c46
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/379052
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Use SGL with contiguous physical buffer metadata as the default PSDT
if NVMe can support SGL.
Change-Id: I99192a8b5408f1b034d5112d77d13639b7558f01
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/378699
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Users don't need to provide nvme debug flag for the error cases.
Change-Id: I00c29e2b8ab470b0233d94acec52b4bec129728c
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.gerrithub.io/376708
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-by: John Kariuki <John.K.Kariuki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This matches the name to the behavior and prepares for addition of a new
log macro for "info" log level.
Change-Id: I94ccd49face4309d3368e399528776ab140748c4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/375833
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
PAGE_SIZE is the host memory page size, which is irrelevant for the NVMe
driver; what we actually care about is the NVMe controller's memory page
size, CC.MPS.
This patch cleans up the uses of PAGE_SIZE in the NVMe driver; the
behavior is still the same in all cases today, since normal NVMe
controllers report a minimum page size of 4096.
Change-Id: I56fce2770862329a9ce25370722f44269234ed46
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/374371
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This also changes the SGL -> PRP case to translate each 4K page from
virtual to physical, in case the buffer is not physically contiguous.
Change-Id: If027f9d656c52c56504f0c64cd4464e16440df63
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/371616
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
For pcie, this just equals the number of SGLs we can fit
into the per-tracker memory.
For rdma, this is just set to 1 for now since nvme_rdma.c
does not support multiple SGEs yet. Once that support is
added, this will change to use MSDBD (Maximum SGL Data Block
Descriptors) instead from the controller identify data.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I34a4c546b5ff46918a296a73ed8cbcc6c9879d5a
Reviewed-on: https://review.gerrithub.io/372358
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Add a new struct spdk_nvme_io_qpair_opts to allow the user to override
controller options on a per-I/O qpair basis.
Existing callers with qprio == 0 can be updated to:
... = spdk_nvme_ctrlr_alloc_io_qpair(ctrlr, NULL, 0);
Callers that need to specify a non-default qprio should be updated to:
struct spdk_nvme_io_qpair_opts opts;
spdk_nvme_ctrlr_get_default_io_qpair_opts(ctrlr, &opts, sizeof(opts));
opts.qprio = SPDK_NVME_QPRIO_...;
... = spdk_nvme_ctrlr_alloc_io_qpair(ctrlr, &opts, sizeof(opts));
Change-Id: I8ac3ea369535cfde759abbe75e1d974b6450a800
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Reviewed-on: https://review.gerrithub.io/369676
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
These commands should be treated as aborted by spec,
so correctly deliver abort notifications when a
qpair is deleted.
Change-Id: I8af47a3f42f5695ef8e1a70813662e69102720b2
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.gerrithub.io/364681
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com>
- rename spdk_malloc_socket to spdk_dma_malloc_socket
- rename spdk_malloc to spdk_dma_malloc
- rename spdk_zmalloc to spdk_dma_zmalloc
- rename spdk_realloc to spdk_dma_realloc
- rename spdk_free to spdk_dma_free
Change-Id: I52a11b7a4243281f9c56f503e826fd7c4a1fd883
Signed-off-by: John Meneghini <johnm@netapp.com>
Reviewed-on: https://review.gerrithub.io/362604
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK Automated Test System <sys_sgsw@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Also avoid an spdk_get_ticks() call in the default
case where a timeout_cb_fn is not defined.
On my Intel(R) Xeon(R) E5-2699 v3 system with an
Intel(R) P3700 SSD, these modifications reduce software
overhead per I/O by 3-5% (as measured by the SPDK
overhead tool).
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I5bb5a87747b15d9e27655fabcd2bc1a40b0b990e
Instead, pass NULL when an ADMIN command times out.
We don't expose the admin queue to the user.
Change-Id: If0768d329a689f6f7c3734c9d419e680d7378ed1
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
For each command that times out, call the timeout
callback one time if the user registered one.
Change-Id: Iaad39a886468e89bef63fe292c5cad1dce97a57c
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
This prevents a deadlock if the user immediately
calls spdk_nvme_detach.
Change-Id: I79f28abe163cbbf184bea907692c44aa4e1c8893
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
The message about the uevent socket is not a fatal error; it just means
that hotplug monitoring will not work.
Change-Id: I29f6a253e96a86420c0fde9e19135f9f1d229bb9
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
remove the unnecessary rte_eal_pci_probe_one() in function
spdk_pci_device_detach(), this could cause error message when we
terminate the application, it will also not make sense try to probe one
device after we detach it, we could call spdk_pci_nvme_device_attach()
instead of spdk_pci_nvme_enumerate() when we have one given device address,
dpdk will try to scan the device and add it back to pci device list then.
Change-Id: I35f5bb412249bb20da57394f0531c10a49691906
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
This can be used for issuing an abort for the timed-out command.
Change-Id: I3c5727fdddc156cd7c8f99afbc3e6da8e73bba56
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Move cb_arg to the first argument to match the other NVMe callback
function signatures.
Change-Id: I4e699c8071dcb7ba4ce3cdb82ee985600208204c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
These were repeated a few different places, so pull them into a common
header file.
Change-Id: Id807fa2cfec0de2e0363aeb081510fb801781985
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This removes one addition from the submission path (negligible, but a
nice side effect), but also opens up the possibility of reporting the
total time an I/O took - since we are always tracking the submission
time anyway, there is no extra cost to report it in the completion
callback.
Change-Id: I7129e7c09d20da8082042a7622d045846461dd9c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The PCIe transport initializes the quirks directly, so the generic hook
to get PCI ID is no longer necessary. This path was dead code.
Change-Id: I25bdaa598db53e4312a264d9d8356d1b416696e5
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The logic to fail queue pairs when the controller is failed should be
handled in the generic code, not in the individual transports.
This also allows nvme_qpair_fail() to be private to nvme_qpair.c.
Change-Id: I6194576dceb35073b9af8847e59314900028637c
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This make sure the qpair failure could be started from upper level application.
Change-Id: I7e04fe36929cc634ddf0078db96fbc40afb38f8c
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
The g_thread_mmio_ctrlr should be not NULL pointer when it enter the
handler function.
Change-Id: I45dba601c672b16e2c6feafd9059bafde0d8f1b4
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
If the user asked for a specific PCI address in spdk_nvme_probe(), we
need to return 1, not 0, for the other PCI addresses that don't match
when enumerating. 0 means to attach the PCI driver, whereas 1 means to
continue enumerating.
With the previous behavior of returning 0, all NVMe devices would be
attached to the DPDK PCI driver, even if the user did not request for
them to be probed, and further calls to spdk_nvme_probe() would not find
any devices.
Change-Id: Ifbbcd7d1abe8ab535b6957855172e66a3e69fbe4
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Currently we use the pci functions provided by DPDK,
it identifies the device by class id related
info but not by pci bdf info, so we can add the filering
by pci_addr in pcie_nvme_enum_cb function.
Change-Id: I5942e98853f00fc10fa6aae5c113517653d1b357
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Since nvme_ns_cmd.c now walks the SGL, some of the test code
needs to also be updated to initialize and return correct values
such as ctrlr->flags and sge_length.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I521213695def35d0897aabf57a0638a6c347632e
Now that the hotplug code is isolated in nvme_pcie.c, it can call the
PCIe transport attach function directly.
Change-Id: I2df3b9168473b537cc9b13367e06d3d3b6fa22be
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The spdk_nvme_qpair::num_entries value is never used in the common code,
so move it to the individual transport qpairs to make it clear that it
is a transport-specific implementation detail.
Change-Id: I5c8f0de4fcd808912ba6d248cf5cee816079fd32
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This function initializes the members of an existing
qpair struct. It doesn't construct one from scratch.
Change-Id: I0b9afac1ad25cfb217efd146702f693c74f5f697
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Make the qpair construct functions private to the transports - it
doesn't need to be called from generic code.
Change-Id: I5f730a4bcf60ce231fe27bc8f4c3c39cb647dd2d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Add a transport callback to return the maximum queue size, and enforce
it in the generic nvme_ctrlr layer.
This allows the user to tell what io_queue_size was actually selected by
the transport via the ctrlr_opts returned during attach_cb.
Change-Id: I8a51332cc01c6655e2a3a171bb92877fe48ea267
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This better describes what the field controls (it does not affect the
admin queue size).
Change-Id: I851ae46fb4ed0fce819af07ae235824e0fc817e6
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
The other simplifications to probe_info and trid made the
trtype argument redundant.
Change-Id: Ie7bea4e2204e690dc4909eeacd065e0722b53272
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>
This can be obtained by parsing traddr into a pci_addr,
then getting a handle to the pci_dev and asking for all
of the pci information.
Change-Id: I1948cbd3ec65611293192ef5558ace19dd444d4c
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Use the standard quirk mechanism to specify which devices
need software assisted striping.
Change-Id: Id8156876a90b4caf9d687637e14c7ad4a66ceda6
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Scanning the transport may result in both new
devices and removed devices, so pass the callback
for both operations.
Change-Id: I6f73dbe6fd7cf61575c354b43f8ae3e2a01e2965
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Simplify the arguments to nvme_transport_ctrlr_scan to take
a transport id that identifies the discovery service (or
NULL to scan PCIe).
Further, separate scan into two functions - scan and attach.
Scan is for scanning an entire bus, attach is for a specific
device.
Change-Id: I464f351a02a04bc5a45096dcf5dc8fc5ac489041
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
It's not the whole transport - it's just an enum for the
type of transport.
Change-Id: Ia435a21792f221ddf50ddf4f0923c6152622eccb
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Change the PCI enumeration API to individual functions per device type
so that only the drivers that are actually in use get linked into the
final executable. All of the common code is still shared internally in
the env_dpdk library.
Change-Id: I2ba83afe59202a510f999a0674e23e60b6581221
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
This version of multi-process support needs to have DPDK 16.11 builtin.
Change-Id: I3352944516f327800b4bd640347afc6127d82ed4
Signed-off-by: GangCao <gang.cao@intel.com>
Each transport should handle its own qpair cleanup internally.
Change-Id: I7dd737be820ea6bad686f4aad7d74044fad58a47
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Let the transport access the controller options during
ctrlr_construct().
Change-Id: I83590c111e75c843685dd9315f0f08416168356d
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Catch SIGBUS and handle it by remapping new memory into the
location where the BAR previously was.
Change-Id: Ie8d00a60a0bbe7f7ec57a5c39c0a63c5d9443206
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
These functions will attach or detach from a PCI device. Attaching
typically means mapping the BAR.
Change-Id: Iaaf59010b8a0366d32ec80bb90c1c277ada7cfe7
Signed-off-by: Cunyin Chang <cunyin.chang@intel.com>
Use the NVMe over Fabrics spec definitions for TRTYPE rather than the
internal library transport type.
Change-Id: Idead559a8f8d95274fc580d10e82033822e6eda8
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
Instead of the next_sge callback returning the physical address
directly, make it return the virtual address and convert to physical
address inside the NVMe library.
This is necessary for NVMe over Fabrics host support, since the RDMA
userspace API requires virtual addresses rather than physical addresses.
It is also more consistent with the normal non-SGL NVMe functions that
already take virtual addresses.
Change-Id: I79a7af64ead987535f6bf3057b2b22aef3171c5b
Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>
