Inference engine is still downloading and building it's own copy of
mkl-dnn, so remove it from DEPENDS.
Signed-off-by: Chin Huat Ang <chin.huat.ang@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
LLD linker is no longer the default for clang-native, so
we can build binaries linking to clang-native using GNU ld.
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
LLD linker is no longer the default for clang-native, so
we can build binaries linking to clang-native using GNU ld.
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
LLD linker is no longer the default for clang-native, so
we can build binaries linking to clang-native using GNU ld.
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Added third party license info:
CMake, 3-clause BSD licenseq
Copyright 2000-2019 Kitware, Inc. and Contributors
See full license text in cmake/Copyright.txt
MathJax, Apache License 2.0
Copyright (c) 2009-2018 The MathJax Consortium
See full license text in doc/assets/mathjax/LICENSE
Boost C++ Libraries, Boost Software License, Version 1.0
Copyright 2005-2014 Daniel James.
See full license text in src/common/LICENSE_1_0
Release highlights can be found here:
https://github.com/intel/mkl-dnn/releases/tag/v1.1
Note:
Starting with this release (version 1.1) the library is renamed to DNNL
https://intel.github.io/mkl-dnn/dev_guide_transition_to_dnnl.htmlhttps://intel.github.io/mkl-dnn/dev_guide_transition_to_v1.html
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
It depends on ace which is marked as incompatible for musl as well
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Add PACKAGECONFIG[python3] for building dldt-inference-engine-python3
package which contains the inference engine python API.
Also tweak recipe to inherit python3native instead of relying on host
python as building the python API requires python3-cython which might
not be available on the host.
Signed-off-by: Chin Huat Ang <chin.huat.ang@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
No need to set LLVM_TARGETS_TO_BUILD here as it
is set by meta-clang layer.
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
tag 'v4.14.146': (60 commits)
Linux 4.14.146
media: technisat-usb2: break out of loop at end of buffer
tcp: Don't dequeue SYN/FIN-segments from write-queue
tcp: Reset send_head when removing skb from write-queue
binfmt_elf: move brk out of mmap when doing direct loader exec
floppy: fix usercopy direction
PCI: kirin: Fix section mismatch warning
iommu/amd: Fix race in increase_address_space()
iommu/amd: Flush old domains in kdump kernel
keys: Fix missing null pointer check in request_key_auth_describe()
x86/hyper-v: Fix overflow bug in fill_gva_list()
x86/uaccess: Don't leak the AC flags into __get_user() argument evaluation
dmaengine: ti: omap-dma: Add cleanup in omap_dma_probe()
dmaengine: ti: dma-crossbar: Fix a memory leak bug
net: seeq: Fix the function used to release some memory in an error handling path
tools/power turbostat: fix buffer overrun
tools/power x86_energy_perf_policy: Fix argument parsing
tools/power x86_energy_perf_policy: Fix "uninitialized variable" warnings at -O2
amd-xgbe: Fix error path in xgbe_mod_init()
perf/x86/amd/ibs: Fix sample bias for dispatched micro-ops
...
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Also updated kmeta
tag 'v4.19.73': (247 commits)
Linux 4.19.73
Merge "USB: Fix configuration selection issues introduced in v4.20.0"
vhost: make sure log_num < in_num
powerpc/tm: Fix restoring FP/VMX facility incorrectly on interrupts
powerpc/tm: Remove msr_tm_active()
PCI: Reset both NVIDIA GPU and HDA in ThinkPad P50 workaround
ext4: unsigned int compared against zero
ext4: fix block validity checks for journal inodes using indirect blocks
ext4: don't perform block validity checks on the journal inode
drm/atomic_helper: Allow DPMS On<->Off changes for unregistered connectors
virtio/s390: fix race on airq_areas[]
drm/i915: Make sure cdclk is high enough for DP audio on VLV/CHV
bcache: fix race in btree_flush_write()
bcache: add comments for mutex_lock(&b->write_lock)
bcache: only clear BTREE_NODE_dirty bit when it is set
NFSv4: Fix delegation state recovery
iio: adc: gyroadc: fix uninitialized return code
mm/migrate.c: initialize pud_entry in migrate_vma()
i2c: at91: fix clk_offset for sama5d2
i2c: at91: disable TXRDY interrupt after sending data
gpio: don't WARN() on NULL descs if gpiolib is disabled
...
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
OE-Core started using Python3 for perf module. But older
kernel such as 4.14 does not have support for python3.
This cause build failure:
| Makefile.config:588: Missing perl devel files. Disabling perl scripting support, please install perl-ExtUtils-Embed/libperl-dev
| Makefile.config:626: Python 3 is not yet supported; please set
| Makefile.config:627: PYTHON and/or PYTHON_CONFIG appropriately.
| Makefile.config:628: If you also have Python 2 installed, then
| Makefile.config:629: try something like:
Backported python3 support, as suggessted here:
https://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/meta/recipes-kernel/perf?id=92469aad50b623afa423c19d82ed2e3c667c5e6a
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Install clDNN to /usr/lib to resolve the following inference engine
error when running with GPU plugin:
[ ERROR ] Failed to create plugin libclDNNPlugin.so for device GPU
Please, check your environment
Cannot load library 'libclDNNPlugin.so': libclDNNPlugin.so: cannot open
shared object file: No such file or directory
/usr/src/debug/dldt-inference-engine/2019r2-r0/git/inference-engine/include/details/os/lin_shared_object_loader.h:36
/usr/src/debug/dldt-inference-engine/2019r2-r0/git/inference-engine/src/inference_engine/ie_core.cpp:277
Signed-off-by: Chin Huat Ang <chin.huat.ang@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
* Release notes:
https://software.intel.com/en-us/articles/OpenVINO-RelNotes
* Enable unit tests to be built and tested using ptest mechanism.
* Include patches from Clear Linux for build fixes.
* Switch to using python3 and threading to using TBB. Switch ENABLE_OPENCV
to off so opencv from system is used.
* Remove do_install and patch Makefiles instead to install libraries correctly.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
This recipe builds the inference engine from opencv/dldt 2019 R1.1
release.
OpenVINO™ toolkit, short for Open Visual Inference and Neural network
Optimization toolkit, provides developers with improved neural network
performance on a variety of Intel® processors and helps further unlock
cost-effective, real-time vision applications.
The toolkit enables deep learning inference and easy heterogeneous
execution across multiple Intel® platforms (CPU, Intel® Processor Graphics)—providing
implementations across cloud architectures to edge device.
For more details, see:
https://01.org/openvinotoolkit
The recipe needs components from meta-oe so move it to
dynamic-layers/openembedded-layer. GPU plugin support needs intel-compute-runtime
which can be built by including clang layer in the mix as well.
CPU and GPU plugins have been sanity tested to work using
classification_sample. Further fine-tuning is still needed to improve
the performance.
Original patch by Anuj Mittal.
Signed-off-by: Chin Huat Ang <chin.huat.ang@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Model Optimizer is a cross-platform command-line tool that facilitates
the transition between the training and deployment environment,
performs static model analysis, and adjusts deep learning models for
optimal execution on end-point target devices.
For more details, see:
https://software.intel.com/en-us/openvino-toolkit/deep-learning-cv
Since the recipe requires bits from meta-python, move this to the
dynamic layers section and add meta-python to BBFILES_DYNAMIC.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Mark 4.14 (it should later be changed to 4.19 after the next LTS has been
merged) as the kernel to be built when using poky-altcfg. It'll help build
and test both the supported kernels.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
The git-lfs content is only used by the samples which we don't use or
install. Set lfs=0 so that git-lfs isn't a requirement on the build host.
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
We'd like to ignore the older tags using year and work week strings like
2018ww19-010806 and use only the ones that have numbers and are in
x.y.z format.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
There is no need to invoke the cmake target explicitly now to have cmake
files installed. Remove the append to do_install doing that.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Also updated KMETA to have drm-bochs feature configurations.
tag 'v4.19.68': (178 commits)
Linux 4.19.68
mmc: sdhci-of-arasan: Do now show error message in case of deffered probe
net/mlx5e: Use flow keys dissector to parse packets for ARFS
net/mlx5e: Only support tx/rx pause setting for port owner
xen/netback: Reset nr_frags before freeing skb
tipc: initialise addr_trail_end when setting node addresses
team: Add vlan tx offload to hw_enc_features
sctp: fix the transport error_count check
sctp: fix memleak in sctp_send_reset_streams
net/packet: fix race in tpacket_snd()
net/mlx4_en: fix a memory leak bug
net: dsa: Check existence of .port_mdb_add callback before calling it
bonding: Add vlan tx offload to hw_enc_features
bnx2x: Fix VF's VLAN reconfiguration in reload.
iommu/amd: Move iommu_init_pci() to .init section
Input: psmouse - fix build error of multiple definition
netfilter: conntrack: Use consistent ct id hash calculation
arm64: ftrace: Ensure module ftrace trampoline is coherent with I-side
dm: disable DISCARD if the underlying storage no longer supports it
drm/i915/cfl: Add a new CFL PCI ID.
...
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Removed patches which are not required anymore.
Updated python version to 3.
In this release, fixed issue/improvements can be
found here:
https://github.com/intel/intel-graphics-compiler/releases/tag/igc-1.0.11
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
With this upgrade, by default -DPREFERRED_LLVM_VERSION="9.0.0" is set.
Changes can be check here:
https://github.com/intel/opencl-clang
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
nGraph Library is an open-source C++ library and runtime/compiler suite
for Deep Learning ecosystem. For more details, see:
https://www.intel.ai/ngraph/
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
OE-core has dropped support for 5.0 and introduced 5.2 linux-yocto
reference kernel. Make changes to remain in sync.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
The library isn't explicitly linking against these libs and loads them
at runtime. Add these to RDEPENDS to make sure these are present.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
The skylake tune introduced on 04510bfa
Currently disables qemu-usermode by default, due to the lack
of support from QEMU to several of the Intel instruction set
extensions introduced within the few latest generations of
CPU architectures (e.g. QEMU does not support avx2)
While there is a good reason to perform the removal of qemu-usermode
from MACHINE_FEATURES, there are several components within the
build system that rely on it for proper compilation and behavior,
for example anything that uses gobject data introspection or even
the components like the chromium web browser require to run a QEMU
for the target architecture to build successfully.
There is no reason why we can't have those components built without
sacrificing the cpu (most) optimizations.
The process I followed on meta-chromebook to enable an optimized build
and whats being upstreamed by this patch is that by doing some
reverse engineering, I was able to figure out which instruction
set extensions are not compatible with QEMU Skylake-Client, by
performing a bit gcc magic from inside QEMU (target) to get the
available optimizations for the native architecture (which is
actually our target in this case).
These are all (not surprisingly) the avx2 extensions, listed as
follows:
-mno-avx
-mno-avx2
-mno-avx512f
-mno-avx512er
-mno-avx512cd
-mno-avx512pf
-mno-avx512dq
-mno-avx512bw
-mno-avx512vl
-mno-avx512ifma
-mno-avx512vbmi
-mno-avx512vbmi2
-mno-avx512vnni
-mno-avx512bitalg
Specifically disabling these manually (for now), allows us to build an
optimized system for the skylake/skylake based architectures (e.g. KabyLake)
while keeping the capability of using qemu-usermode, as a side note GCC
does show more unavailable optimization tunes, (hence why there might be
some warnings), but getting rid of these specifically seems enough to make
it run happily in qemu-usermode.
Basically the MACHINE_FEATURES variable is able to dictate how we will tune
the build for our device, if qemu-usermode is present, TUNE_CCARGS will
expand as follows:
TUNE_CCARGS=" -m64 -march=skylake -mtune=skylake -mno-avx -mno-avx2
-mno-avx512f -mno-avx512er -mno-avx512cd -mno-avx512pf -mno-avx512dq
-mno-avx512bw -mno-avx512vl -mno-avx512ifma -mno-avx512vbmi -mno-avx512vbmi2
-mno-avx512vnni -mno-avx512bitalg -mfpmath=sse"
Whats this means is that the build will be tuned for skylake architectures,
enabling all possible extensions, (MOVBE, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1,
SSE4.2, POPCNT, AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC and XSAVES) minus the ones QEMU is unable
to run, which will result in an illegal instruction error, notice the tune is
kept as skylake.
Whilst, if qemu-usermode is not found on MACHINE_FEATURES, TUNE_CCARGS will
expand to:
TUNE_CCARGS=" -m64 -march=skylake -mtune=generic -mavx2 -mfpmath=sse"
Which is exactly what its set to with the current tune, so this patch should
be harmless, and only extend current functionality, we will still keep
qemu-usermode disabled by default (hence enabling avx2) but the user would
have the capability to decide how the skylake tune will be handled.
As the GCC manual states, we should really try to avoid using mtune=generic
when possible, and this patch allows us to do so [1].
This patch also addresses a problem on which the current skylake tune includes
tune-core2.inc instead of tune-corei7.inc to get the list of AVAILTUNES
and PACKAGE_EXTRA_ARCHS.
Right now, AVAILTUNES are set as follows:
AVAILTUNES=" x86 x86-64 x86-64-x32 i586 i686 core2-32 core2-64 core2-64-x32
skylake-64"
Where the proper set should be (after this patch):
AVAILTUNES=" x86 x86-64 x86-64-x32 i586 i686 core2-32 core2-64 core2-64-x32
corei7-32 corei7-64 corei7-64-x32 skylake-64"
When (if), QEMU gains support for the AVX2 instruction set extensions these
can be easily removed to provide full support for qemu-usermode.
[1] https://gcc.gnu.org/onlinedocs/gcc-9.1.0/gcc/x86-Options.html
Signed-off-by: Alejandro Enedino Hernandez Samaniego <aehs29@gmail.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
The latest version is 2.27.0, not 2.0.27. Also make sure that we are
pointing to the release commit.
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
For LICENSE updates, pleaes check:
https://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/backport-iwlwifi.git/commit/COPYING?h=release/core40&id=8307aa7b67e71e4d4abe1693107eba63cfd3702b
More information about this release
can be found here:
https://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/backport-iwlwifi.git/log/?h=release/core45
This update is tested on NUC7 having 'Intel(R) Dual Band Wireless AC 8265' wireless module.
Supported firmwares/devices:
iwlwifi-7265D-29.ucode
iwlwifi-7265-17.ucode
iwlwifi-3168-29.ucode
iwlwifi-3160-17.ucode
iwlwifi-7260-17.ucode
iwlwifi-8265-36.ucode
iwlwifi-8000C-36.ucode
iwlwifi-9260-th-b0-jf-b0-46.ucode
iwlwifi-9000-pu-b0-jf-b0-46.ucode
iwlwifi-ty-a0-gf-a0-48.ucode
iwlwifi-so-a0-gf-a0-48.ucode
iwlwifi-so-a0-hr-b0-48.ucode
iwlwifi-so-a0-jf-b0-48.ucode
iwlwifi-cc-a0-48.ucode
iwlwifi-QuQnj-b0-jf-b0-48.ucode
iwlwifi-QuZ-a0-jf-b0-48.ucode
iwlwifi-QuZ-a0-hr-b0-48.ucode
iwlwifi-Qu-b0-jf-b0-48.ucode
iwlwifi-Qu-c0-hr-b0-48.ucode
iwlwifi-QuQnj-a0-hr-a0-48.ucode
iwlwifi-QuQnj-b0-hr-b0-48.ucode
iwlwifi-Qu-b0-hr-b0-48.ucode
iwlwifi-QuQnj-f0-hr-a0-48.ucode
iwlwifi-Qu-a0-jf-b0-48.ucode
iwlwifi-Qu-a0-hr-a0-48.ucode
Most of the above firmwares are provided by below recipe version (20190815):
https://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/meta/recipes-kernel/linux-firmware/linux-firmware_20190815.bb
SRCREV = "07b925b450bfb4cf3e141c612ec5b104658cd020"
SRC_URI = "git://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git"
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Add 64-bit new machine (intel-skylake-64) with -march=skylake
and avx2 instruction-set set up.
We do see a qemu-usermode failure at build time, on setup of
avx2 instruction-set as QEMU does not support AVX instruction
set.
Check this:
https://bugs.launchpad.net/qemu/+bug/1818075
So to bypass this issue disabling qemu-usermode for intel-skylake-64
machine.
Due to above limitation and in order to not affecting existing
machines, this new machine is being proposed to add.
A quick performance comparision between intel-corei7-64 vs intel-skylake-64 machines
Measurements are in time, taken by the benchmark tests. Less is better.
Test/Benchmark intel-corei7-64 intel-skylake-64
(1) CppPerformanceBenchmarks (Test: Math Library)
(1st) 5m 15.70s 4m 36.39s
(2nd) 5m 16.37s 4m 36.51s
(3rd) 5m 15.54s 4m 37.80s
CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks.
(2) AOBench
(1st) 0m 35.07s 0m 28.74s
(2nd) 0m 34.90s 0m 28.72s
(3rd) 0m 34.85s 0m 28.89s
AOBench is a lightweight ambient occlusion renderer, written in C. The
test profile is using a size of 2048 x 2048.
(3) C-Ray
(1st) 320 seconds 232 seconds
(2nd) 320 seconds 232 seconds
(3rd) 321 seconds 232 seconds
C-Ray, a simple raytracer designed to test the floating-point CPU performance.
For this patch, 'bitbake world' gets successfully built with latest poky master.
Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
Switch the driver configuration MEDIA_BUILD_FATAL_WARNINGS to OFF so
-Werror is not used. Also patch out the use of
-Werror=implicit-function-declaration to prevent errors when building
with gcc9.
| cc1plus: error: '-Werror=' argument
'-Werror=implicit-function-declaration' is not valid for C++ [-Werror]
Signed-off-by: Anuj Mittal <anuj.mittal@intel.com>
