Bit 1 in the CMIC of the Identify Controller Data Structure specifies
if the NVM subsystem may have multiple controllers or not.
However, multi_host indicated a particular use case such that the NVM
subsystem is used by multiple hosts.
multi_ctrlr will be more appropriate.
Signed-off-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Change-Id: I0246096a5cc44721aeff3ff6f96473a2abe11964
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/8719
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Adding iov to the spdk_bdev_zcopy_start function enable spdk_bdev_zcopy_start to
be used by transport layers as the iov is owned by the transport command
Signed-off-by: matthewb <matthew.burbridge@hpe.com>
Change-Id: I6d2be7f49566048bf25b7711ada8d2fb49fea6ee
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6816
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Michael Haeuptle <michaelhaeuptle@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This is useful for applications even if they elect not to use the SPDK
event framework.
This doesn't shift everything in one go - just the subsystem
initialization logic. Configuration file loading also needs to move
in a separate patch later.
Change-Id: Id419df1045442d416650ed90e5ee78adfdd623d7
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6641
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
This include isn't needed in queue_extras.h itself.
There were a few places that were implicitly
depending on this include, so fix those to include
util.h explicitly.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia962ae5a4403ee8ae15f3106d0d5e7d7412a4535
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/7172
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: <dongx.yi@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
These tests compile the driver both as a shared object and statically
linking it to the identify application. Additionally, in both
configurations, the app is used to list all available NVMe controllers,
as well as print information about each one indivdually.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I127caf08acad11241bf685b392617ab4b810226b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6680
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Replaced the "_tc[0-9]+" suffix with a more descriptive title describing
the test case being executed.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Id1c2f6dcf155d8ae99dc490983ab4bb1451bfefd
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6764
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Removed the PCI_ALLOWED="NONE" line from setup.sh calls to bind to all
available PCI devices. It'll allow the out-of-tree NVMe driver tests
from the following patch to execute properly, as they require NVMe
devices to be attachable from SPDK.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I56e89a6b853f02343803bb6ec704ea3c0a8bd12f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6679
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
The application will now print some of the fields from the identify
controller data.
The code has been copied from `examples/nvme/identify`, but, for the
sake of simplicity, trimmed down to printing reduced number of fields.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I838c47deffb0b877344f3cad0e88b6aca19790ea
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6678
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Now that it's possible to both submit the identify controller command
and process its completion, the initialization flow has been updated to
issue that command and update controller's identify data (which can be
retrieved via nvme_ctrlr_get_data()).
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Iee0e85f431275a5e6f1767db1d807de7fba6cdcc
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6677
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Added function that check the completion queue for completed commands,
executes their callbacks, and puts the associated requests back onto the
free request queue.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I0f04c0d173a7058d4d4f7e59e573ce48130ff024
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6676
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
This patch adds code required for sending identify controller admin
command. It means allocating an NVMe request, filling spdk_nvme_cmd
structure, and submitting the command by writing the command to the
submission queue and ringing the doorbell.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I69934213a350df03852860eeaeadc2a456c7673c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6675
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
This function allows the user to retrieve the controller's identify
data. Currently, that buffer is zeroed, but it'll be populated by the
identify command.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I91a99feef25ecf94c43cf144c12ac3c541c76cd9
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6674
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Added a simple controller enablement state machine based on the CC.EN
and CSTS.RDY bits. The admin queue registers are also filled during
this process, so it's now possible to send admin requests.
To simplify the code, there are no timeouts for a controller to
transition from a specific state to the next one or for the whole
initialization process. This means that if a controller gets stuck, the
code will hang indefinitely too.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I93f5a5931d7b24780da242e601dcdf2bec5f6552
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6673
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
After enumerating and attaching NVMe controllers, they're now
initialized at the end of nvme_probe()/nvme_connect(). For now, they're
immediately marked as initialized, but subsequent patches will replace
it with an actual initialization.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I22137bb10e871c7e79c28053c8ec98a835e11147
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6672
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This patch adds NVMe submission/completion queue pair definitions.
These definitions are required to keep track of outstanding NVMe
requests. The admin queue pair is now instantiated with the minumum
number of entries (2).
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I2ced3ce7d210408d66cc17de1e66d86b1a1dbf79
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6671
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
These values are needed for managing a submission/completion queue pair,
which will be added in the subsequent patch.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I80ac0d607160f06a13014b7dea95ae8172290aee
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6670
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
It allows the controller to issue memory read/writes (the bus master
enable bit) and disables the ability to generate INTx interrupts which
won't be serviced.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I3b041f1ea7c2bc275b609afcc3d1e4f655aee4c5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6669
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Added getters/setters providing access to several of the NVMe
controller's registers. Only the registers that are needed for the
initialization are implemented.
For now all of them are unused, so they're marked as external to avoid
the -Wunused-function warnings. The subsequent patches will make use of
them and mark as static appropriately.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I7012583f74e87720f6915afca69474ad1bb1e377
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6668
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
The controller's memory register space (located in the first BAR) is now
mapped. The functions for accessing individual registers from this area
will be added in the following patch.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Ie5f88079a46152ba8d68e534d5e4c0c2bef84ef3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6667
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
The identify app can now be used in two modes:
- without any parameters it'll attach to all available NVMe controllers
and print each one out,
- with a single parameter specifying the BDF address of the controller
it'll only attach and print out information about that controller.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I02c7a8a072f1db5fdfd428a5ab84163f26338a09
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6666
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Added interface, nvme_connect(), for connecting an NVMe device specified
by a BDF address. Similarly to spdk_nvme_connect(), it returns a
pointer to an NVMe controller representing requested device.
Under the hood, it uses spdk_nvme_pci_device_attach() to attach the
PCI device.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I15c7cabae07539cb97aba810385d26f223abfd6d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6665
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
The NVMe devices can now be enumerated, attached, and detached. To
simplify the driver, the probing step has been omitted - all available
controllers are attached and need to be detached later.
The driver registers itself as a PCI driver via a call to
SPDK_PCI_DRIVER_REGISTER() and then uses spdk_pci_enumerate() to
enumerate available NVMe devices and attach them.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Id03e2f4365f4f7ca98178be70278d0c4b7b34b26
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6664
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This patch introduces initial definitions for a stripped down version of
an out-of-tree NVMe driver. It's purpose is to showcase SPDK's
interfaces for writing user-space drivers for various types of devices.
The choice of NVMe as the example is based on the fact that it can be
emulated by QEMU and the code can be borrowed from SPDK's regular NVMe
driver.
This driver will only provide the most basic functionalities (e.g.
device probing, controller initialization, only admin queue support,
etc.) and won't have support for any device quirks. Therefore, it's
only intented to be used with emulated devices.
In addition, an application utilizing the driver to list all available
NVMe controllers and print their identify data will be added. It'll
be a very basic, stripped down version of `examples/nvme/identify`.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I67c748aabf75a37ca72dfb74301a610f7c4ae2bc
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6663
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
- it shows nice feature and usage of pkg-config
- it also prevents from duplicate symbols issue for static build in
case listed libs have a common dependencies
Signed-off-by: Jacek Kalwas <jacek.kalwas@intel.com>
Change-Id: I1f470d202ea7efe03e354dca472fd50e97bca747
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6406
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
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: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Static compilation for external_code tests uses SYS_LIBS
that now includes libvfio-user. As such the COMMON_CFLAGS
need to point to the build directory.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I6f3aa670a62445088535eacdfa1483bfacd5ce7b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6020
Community-CI: Broadcom CI
Reviewed-by: Karol Latecki <karol.latecki@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Users can now generate the necessary linker args for their
own applications using something like:
PKG_CONFIG_PATH=build/lib/pkgconfig pkg-config --libs spdk_nvme
Dependencies between libraries are included in the generated
.pc files, so the user only needs to pass the top-level subsystems
or individual SPDK libraries they are using in their application.
Modules will automatically be added to the output if the associated
library is specified. For example, specifying "spdk_bdev" will include
the libraries not only for spdk_bdev, but also all of the bdev modules.
Users still need to supply the -Wl,--no-as-needed or -Wl,--whole-archive
flags. They cannot be added to the .pc files without increasing the length
of the argument string by a factor of 15x to 20x.
Modify the test/external_code/hello_world Makefile to use pkg-config to
ensure this gets tested at some level in our autotest environment.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie48a75f11969d5d775d514cf10bcb82d197eabfd
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/4371
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
-pthread is what SPDK in-tree Makefiles use, so use that
in the external_code Makefile too. -pthread can be
used both on the compilation and link stages.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I8ef702b8fbde825a4191d961738ebc605b0aa4c5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/5675
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
The "external" module that we link against is essentially
a copy of the passthru bdev module in the modules/bdev
directory. But let's not use the same names for the
externally visible functions from this "external" module,
so that we don't get link errors if we happen to link
in both modules into one application.
This will help in a future patch where we use pkg-config
files to generate the list of the bdev modules.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I835c7e7ee8902f6829657ec6a3cf40cd802a1357
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/5674
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Our examples here show how to link in the DPDK environment
into an SPDK application. That includes both spdk_env_dpdk
library and the DPDK libraries. So put spdk_env_dpdk in
the DPDK_LIB variable in the Makefile.
This will also simplify a future patch where we will
use pkg-config in this Makefile instead.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I2c389eaa9900dbe5f77ccee536bd4edc0a54cccc
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/5673
Community-CI: Broadcom CI
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
test/external_make wants to always default to
the autotest_common.sh HUGEMEM default which is
8192MB on Linux. But it is useful to be able to
override this to run this test in VMs with less
memory.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I23a40c6c14e35cffefbaf4fa993ebd28d1b0054c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/5672
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Similarly replace PCI_BLACKLIST with PCI_BLOCKED.
Use of ALLOWED/BLOCKED matches similar changes made
in DPDK.
While here, replace use of term "blacklist" with "blocked"
in one of the nvme perf scripts. The usage there was
associated with how devices are blocked by using the
environment variables that are changed by this patch.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I720d99118ba5e050f436612c9fd415db44294a63
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/5275
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
There are many duplicated codes about sending response for writing bool result.
That we need a function to do this.
Then we can reduce many codes.
Signed-off-by: yidong0635 <dongx.yi@intel.com>
Change-Id: Ic439111b1e9ca1013f8c657ab925f0c27a7be699
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/5033
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This test script is based around only a strict
subset of the SPDK configuration options. If we don't
clean before building, we could still have libraries
in build/lib from previous tests that the external_code
Makefiles do not account for when linking libspdk.so,
which will include *all* libraries in build/lib.
Fixes issue #1662.
Fixes issue #1665.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I525ea1740313048fc44f26b5b0db9ab38161725e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/4970
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: <dongx.yi@intel.com>
rte_hash depends on rte_rcu starting in upcoming
DPDK 20.11 release. rte_rcu was only added in
DPDK 19.05 release, so we need to check if it
exists before linking it.
Fixes issue #1661.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I7e343c6f964b03cc62484b57803a3bad00f80288
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/4947
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Signed-off-by: Maciej Szwed <maciej.szwed@intel.com>
Change-Id: Ide0ac38497bc500886e26fd219cb120113896966
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/4152
Tested-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The files for external_code tests were copies of the same files
in SPDK. Intention of external_code tests is to verify
that nothing breaks in regards to building process.
The copies were getting out of sync with functional changes.
So to address this:
- hello_bdev app is now symlinked
- vbdev_passthrough is copied from respective files
- RPC and name of the module were changed to not conflict with
the original passthrough module
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: Idb83b8a06dea7143505d2062b23bfb35c86e3274
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/4804
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Get rid of WITH_DPDK_DIR and SPDK_RUN_INSTALLED_DPDK,
introduce SPDK_RUN_EXTERNAL_DPDK which can point to
a DPDK dir. It's an empty string by default.
Change-Id: Iff2b3773a4614db07f4196165087a79472e02b9a
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/867
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Makefile support in DPDK was deprecated and will be removed soon,
so switch to the officially supported way of building DPDK -
with meson and ninja. Two new tools. Basically, our Makefiles
will invoke meson+ninja for DPDK, no other SPDK components are
affected.
Apparently DPDK wanted to move away from an octopus-like config
system and the ideology behind meson configuration is simple now:
build everything by default. Some PMDs can be explicitly disabled
with meson command line, but all libraries (both static and shared
versions) and test apps are built unconditionally.
How long does it take to build minimal DPDK with meson? Too much.
On my machine half of the total build time is spent on libraries
we don't need at all. (I have some hacks up my sleeve to disable
building those libraries - see the subsequent patch.) As for the
official way of building a minimal DPDK, there was a patch [1]
on dpdk mailing list to introduce more specific configuration,
but it was rejected:
> We talked about this a few times in the past, and it was actually one
> of the design goals to _avoid_ replicating the octopus-like config
> system of the makefiles. That's because it makes the test matrix
> insanely complicated, not to mention the harm to user friendliness,
> among other things.
>
> If someone doesn't want to use a PMD, they can just avoid installing it
> - it's simple enough.
>
> Sorry, but from me it's a very strong NACK.
Let's not follow that direction, hack the DPDK build system instead.
As for advantages of meson+ninja over Makefiles? I can't find any.
It's another build system that does a lot for you with some functions,
magic options, and a built-in dependency system. It seems nice if you know
the syntax, but it's another component that you need to learn, debug,
and possibly find bugs in (there's a lot of github issues open for meson).
I would compare it to CMake.
As for changes in this patch: rather that explicitly disabling
PMDs we don't need, specify a list of PMDs we do need and disable
everything else found in ./dpdk/drivers/*. This way we won't have
to disable the new PMDs as they're added to DPDK.
Meson configuration also sets RTE_EAL_PMD_PATH #define to a valid directory
with built PMD shared libs. When it's set, DPDK dynamically loads all shared
libraries inside. The drivers there depend on DPDK shared libs and fail to
load in static SPDK builds, so we disable them altogether by unsetting
RTE_EAL_PMD_PATH in the meson-generated config file - just like
DPDK Makefiles did. EAL checks for RTE_EAL_PMD_PATH being empty and skips
loading any external PMDs then. We do it for both static and shared libs.
We specify all PMDs at build time for now, so there's just no need to load
them dynamically.
We have three more hacks in our submodule:
* disable building dpdk apps by commenting-out a line in dpdk/meson.build
* disable building unnecessary libs (build everything that spdk *may*
need)
* build isa-l compress pmd with `-L[...] -lisal`. DPDK expects to find
libisal with pkg-config. We don't want to prepare a pkg-config file,
so comment-out a failing check in another meson.build file and provide
isa-l through CFLAGS and LDFLAGS.
We also need to make some changes to our test/external_code. First of
all, -ldpdk is no more. Meson build generates a pkg-config file with all
libs, but we'll switch to it in a separate patch - for now just specify
all -lrte_ libs one by one. -Wl,--no-as-needed has to be added to some
test cases, otherwise rte_mempool_ring isn't loaded. We don't use any
APIs from this library, it only has a static constructor that provides
a few callbacks used by rte_mempool_create(). Also, since DPDK now builds
both static and shared libraries, we need to add -Wl,-Bstatic to force
using static libswhere required. It's only needed for DPDK libs, but we
use it for SPDK libs as well since there's no harm.
As for performance:
$ ./configure --enable-debug --with-crypto --with-reduce
$ time make -j40 -C dpdkbuild all
with meson:
real 0m8.287s
user 1m7.983s
sys 0m10.548s
before, with the old DPDK makefiles:
real 0m20.232s
user 0m55.921s
sys 0m16.491s
The subsequent builds are much faster too:
$ time make -j40 -C dpdkbuild all
meson:
real 0m0.876s
user 0m0.663s
sys 0m0.217s
makefiles:
real 0m10.150s
user 0m11.740s
sys 0m6.772s
[1] http://inbox.dpdk.org/dev/1a07d1cd59d84dce84e56c10fdabf5e5504560a6.camel@debian.org/
Change-Id: Ic65db563014100bafb12e61ee0530cc2ae64401d
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1440
Community-CI: Mellanox Build Bot
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>
With update to DPDK 20.05 EAL requires rte_telemetry,
external make tests were missing specifying this lib.
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I9f5eaea46b47da48fc81b4fff0199cd75ff6af78
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/3483
Community-CI: Mellanox Build Bot
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: Jim Harris <james.r.harris@intel.com>
Just set it once instead of doing it over and over.
This cleans up the code a lot.
Signed-off-by: Seth Howell <seth.howell@intel.com>
Change-Id: I91c6ea8c02d964693909b9c53795b0f1ee09c90f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/2979
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Fixes#1382
System memory gets fragmented over time and repeated hugepage
allocations fail, giving us intermittent failures on CI. Linux
allocates those pages only when requested hugepage count is bigger
than current hugepage count, so in order to prevent any additional
allocations just stick to a single count all the time.
autotest_common.sh already exports HUGEMEM=8G for Linux systems
and 2G for BSD ones. `setup.sh reset` doesn't touch the hugepages,
so practically only the first `setup.sh` will allocate memory now.
The two setup.sh calls are run with sudo, so they need
HUGEMEM="$HUGEMEM" to work.
Change-Id: I48b4a720bad4e1da14e78313b4db0eb60a177c38
Signed-off-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/2374
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: Seth Howell <seth.howell@intel.com>
This makes sure that we will have the hugepages we need
even on systems that don't have them configured by default.
Signed-off-by: Seth Howell <seth.howell@intel.com>
Change-Id: I9a0d82e0abf6363c1c037732b543c05e8931f97e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/2141
Community-CI: Mellanox Build Bot
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: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This test helps ensure that nothing about the way we build our shared
objects changes in a way that breaks external applications or bdev
modules trying to link to SPDK.
Change-Id: I45ac36a2afc873b04eb237f4e633a9d84066737c
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/435
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI
