tensorflow2-lite-2.6.2-bp154.1.26

- Remove more python dependencies for tensorflow2-lite

- tensorflow2-lite version does not need all the python dependencies
  listed for tensorflow2

- Leap 15.x / Backports: Do not build non-Lite versions since python3-numpy
  and python3-scipy are too old for Keras/TF2

- updated to 2.6.2 which is bug fix release which just fixes an issue where
  keras, tensorflow_estimator and tensorboard were missing proper upper bounds
  and resulted in broken installs after Keras 2.7 release for all packages in
  TensorFlow ecosystem
- Fixes from 2.6.1 (boo#1192447):
  * Fixes a code injection issue in saved_model_cli (CVE-2021-41228)
  * Fixes a vulnerability due to use of uninitialized value in Tensorflow (CVE-2021-41225)
  * Fixes a heap OOB in FusedBatchNorm kernels (CVE-2021-41223)
  * Fixes an arbitrary memory read in ImmutableConst (CVE-2021-41227)
  * Fixes a heap OOB in SparseBinCount (CVE-2021-41226)
  * Fixes a heap OOB in SparseFillEmptyRows (CVE-2021-41224)
  * Fixes a segfault due to negative splits in SplitV (CVE-2021-41222)
  * Fixes segfaults and vulnerabilities caused by accesses to invalid memory
    during shape inference in Cudnn* ops (CVE-2021-41221)
  * Fixes a null pointer exception when Exit node is not preceded by Enter op (CVE-2021-41217)
  * Fixes an integer division by 0 in tf.raw_ops.AllToAll (CVE-2021-41218)
  * Fixes a use after free and a memory leak in CollectiveReduceV2 (CVE-2021-41220)
  * Fixes an undefined behavior via nullptr reference binding in sparse matrix
    multiplication (CVE-2021-41219)
  * Fixes a heap buffer overflow in Transpose (CVE-2021-41216)
  * Prevents deadlocks arising from mutually recursive tf.function objects (CVE-2021-41213)
  * Fixes a null pointer exception in DeserializeSparse (CVE-2021-41215)
  * Fixes an undefined behavior arising from reference binding to nullptr in
    tf.ragged.cross (CVE-2021-41214)
  * Fixes a heap OOB read in tf.ragged.cross (CVE-2021-41212)
  * Fixes a heap OOB in shape inference for QuantizeV2 (CVE-2021-41211)
  * Fixes a heap OOB read in all tf.raw_ops.QuantizeAndDequantizeV* ops (CVE-2021-41205)
  * Fixes an FPE in ParallelConcat (CVE-2021-41207)
  * Fixes FPE issues in convolutions with zero size filters (CVE-2021-41209)
  * Fixes a heap OOB read in tf.raw_ops.SparseCountSparseOutput (CVE-2021-41210)
  * Fixes vulnerabilities caused by incomplete validation in boosted trees code (CVE-2021-41208)
  * Fixes vulnerabilities caused by incomplete validation of shapes in multiple TF ops (CVE-2021-41206)
  * Fixes a segfault produced while copying constant resource tensor (CVE-2021-41204)
  * Fixes a vulnerability caused by unitialized access in EinsumHelper::ParseEquation (CVE-2021-41201)
  * Fixes several vulnerabilities and segfaults caused by missing validation
    during checkpoint loading (CVE-2021-41203)
  * Fixes an overflow producing a crash in tf.range (CVE-2021-41202)
  * Fixes an overflow producing a crash in tf.image.resize when size is large (CVE-2021-41199)
  * Fixes an overflow producing a crash in tf.tile when tiling tensor is large (CVE-2021-41198)
  * Fixes a vulnerability produced due to incomplete validation in
    tf.summary.create_file_writer (CVE-2021-41200)
  * Fixes multiple crashes due to overflow and CHECK-fail in ops with large
    tensor shapes (CVE-2021-41197)
  * Fixes a crash in max_pool3d when size argument is 0 or negative (CVE-2021-41196)
  * Fixes a crash in tf.math.segment_* operations (CVE-2021-41195)

- Make sure tensorflow/core/public/version.h is installed in the
  'lite' version (bsc#1191805).

- Add missing python requirements -- boo#1190856

- Limit BuildRequires for bazel-skylib-source to versions >= 1.0.3.

- Update to 2.6.0
  Major changes are:
  * Keras been split into a separate PIP package (keras), and its
    code has been moved to the GitHub repositorykeras-team/keras.
    The API endpoints for tf.keras stay unchanged, but are now
    backed by the keras PIP package. The existing code in
    tensorflow/python/keras is a staled copy and will be removed in
    future release (2.7). Please remove any imports to
    tensorflow.python.keras and replace them with public tf.keras
    API instead.
  * tf.train.experimental.enable_mixed_precision_graph_rewrite is
    removed, as the API only works in graph mode and is not
    customizable. The function is still accessible under
    tf.compat.v1.mixed_precision.enable_mixed_precision_graph_rewrite,
    but it is recommended to use the Keras mixed precision API
    instead.
  * tf.lite: Remove experimental.nn.dynamic_rnn,
    experimental.nn.TfLiteRNNCell and
    experimental.nn.TfLiteLSTMCell since they're no
    longer supported. It's recommended to just use keras lstm
    instead.
  * tf.keras: The methods Model.to_yaml() and
    keras.models.model_from_yaml have been replaced to raise a
    RuntimeError as they can be abused to cause arbitrary code
    execution. It is recommended to use JSON serialization instead
    of YAML, or, a better alternative, serialize to H5.
- Major changes from 2.5.x:
  * Support for Python3.9 has been added.
  * The TF_CPP_MIN_VLOG_LEVEL environment variable has been renamed
    to to TF_CPP_MAX_VLOG_LEVEL which correctly describes its
    effect.
- Fixed multiple CVEs (boo#1189423):
  * CVE-2021-37635
  * CVE-2021-37636
  * CVE-2021-37637
  * CVE-2021-37638
  * CVE-2021-37639
  * CVE-2021-37640
  * CVE-2021-37642
  * CVE-2021-37641
  * CVE-2021-37644
  * CVE-2021-37643
  * CVE-2021-37645
  * CVE-2021-37646
  * CVE-2021-37647
  * CVE-2021-37648
  * CVE-2021-37649
  * CVE-2021-37650
  * CVE-2021-37651
  * CVE-2021-37652
  * CVE-2021-37653
  * CVE-2021-37654
  * CVE-2021-37655
  * CVE-2021-37656
  * CVE-2021-37657
  * CVE-2021-37658
  * CVE-2021-37659
  * CVE-2021-37660
  * CVE-2021-37661
  * CVE-2021-37662
  * CVE-2021-37664
  * CVE-2021-37663
  * CVE-2021-37665
  * CVE-2021-37666
  * CVE-2021-37667
  * CVE-2021-37668
  * CVE-2021-37669
  * CVE-2021-37670
  * CVE-2021-37671
  * CVE-2021-37672
  * CVE-2021-37673
  * CVE-2021-37674
  * CVE-2021-37676
  * CVE-2021-37675
  * CVE-2021-37677
  * CVE-2021-37678
  * CVE-2021-37679
  * CVE-2021-37680
  * CVE-2021-37681
  * CVE-2021-37682
  * CVE-2021-37683
  * CVE-2021-37684
  * CVE-2021-37686
  * CVE-2021-37685
  * CVE-2021-37687
  * CVE-2021-37688
  * CVE-2021-37689
  * CVE-2021-37691
  * CVE-2021-37692
  * CVE-2021-37690
- Updated sources:
  * abseil-cpp.tar.gz
  * cpuinfo.zip
  * dill-0.3.2.zip
  * eigen.tar.gz
  * google-cloud-cpp.tar.gz
  * libxsmm_1.14.tar.gz
  * llvm.tar.gz
  * oneDNN.tar.gz
  * rules_cc.tar.gz
  * rules_closure.tar.gz
  * rules_docker-0.18.0.tar.gz
  * ruy.zip
  * tblib-1.7.0.tar.gz
- Added sources:
  * ComputeLibrary.tar.gz
  * oneDNN-v2.3-rc2.tar.gz
  * platforms-0.0.2.tar.gz
  * rules_proto.tar.gz
  * tf_runtime.tar.gz
  * tf_toolchains.tar.gz
- Removed sources:
  * kafka-v0.11.5.tar.gz
- Add "tensorflow-2.6.0" prefix to existing patches to indicate
  that patches are likely to be only applicable to a specific
  version.
  * fix-lite.patch -> tensorflow-2.6.0-fix-lite.patch
  * numpy-tensor-small.patch -> tensorflow-2.6.0-numpy-tensor-small.patch
  * removed-clog-build-as-included-in-cpuinfo.patch -> tensorflow-2.6.0-removed-clog-build-as-included-in-cpuinfo.patch
  * removed-external-toolchains.patch -> tensorflow-2.6.0-removed-external-toolchains.patch
  * remove-weakref.patch -> tensorflow-2.6.0-remove-weakref.patch
  * tf-keras-hdf5-3.patch -> tensorflow-2.6.0-tf-keras-hdf5-3.patch
- Rebase all existing patches.
- Add tensorflow-2.6.0-compile-with-protobuf-3.16.patch to fix
  build error with protobuf >= 3.16.0. (boo#1186860)
  (https://github.com/protocolbuffers/protobuf/pull/8354)
- Update bazel version requirement to 3.7.2.
- Drop pcre-devel build requirement as it is not used anymore.
- Drop --incompatible_no_support_tools_in_action_inputs=false as it
  is removed in bazel >= 3.6.

- Update _constraints to use host with 'asimdrdm' cpu flag to avoid
  slow CPU and be scheduled on faster systems

- Update to version 2.4.1
  * Bugfixes
  * Drops requirement of AVX2

- Don't BuildRequire keras_applications. Tensorflow provides it
  itself: https://github.com/tensorflow/tensorflow/commit/23c3bdaa
- These were discovered by Keras test suite:
  * add numpy-tensor-small.patch for Numpy >= 1.20
    gh#tensorflow/tensorflow#47691
  * add tf-keras-hdf5-3.patch for hdf5 >= 3.0
    gh#tensorflow/tensorflow#44467

- Generate and install pkgconfig files for tensorflow-lite and
  tensorflow (non-hpc)

- Do not install bazel tools to build Lite version. This will allow
  to build for armv7 where bazel 3.x is not available - boo#1178564

- Fix packaging for libiomp5

- build verbose to not fail on the obs stall detection

- libiomp5 is x86_64 only

- Fix build on aarch64 and on hpc variants

- Updated to version 2.4.0 which fixes several bugs (bsc#1173128)
  (bsc#1173314)i, (bsc#1179455) and (bsc#1178287)
- Security fixes for CVE-2020-26266, CVE-2020-26267, CVE-2020-26268,
  CVE-2020-26270 and CVE-2020-26271
- updated sources:
  * abseil-cpp.tar.gz
  * eigen.tar.gz
  * gemmlowp.zip
  * googleapis.zip
  * llvm.tar.gz
- added sources:
  * DouraFFT.tar.gz
  * cpuinfo.zip
  * dill-0.3.1.1.tar.gz (aarch64 only)
  * dlpack.tar.gz
  * oneDNN.tar.gz
  * openmp-10.0.1.src.tar.xz
  * python-license-astunparse
  * re2.tar.gz
  * rules_android.zip
  * rules_cc.tar.gz
  * rules_docker-0.15.0.tar.gz
  * ruy.zip
  * sobol_data.tar.gz
  * tblib-1.3.2.tar.gz (aarch64 only)
  * typing_extensions-3.7.4.2.tar.gz (aarch64 only)
- removed sources:
  * boring_ssl.tar.gz
  * cub_1.8.0.zip
  * mkl-v0.21.2.tar.gz
  * pybind11-v2.3.0.tar.gz
  * right-json-location.patch
  * rules_docker.tar.gz
- removed patches:
  * Provide-overload-to-cope-with-const-ness-change-of-N.patch
  * fix-google-absl-memory.patch
  * json-feature-name.patch
  * libjpeg_turbo-name.patch
  * removed-docker-tools.patch
- added patches:
  * removed-clog-build-as-included-in-cpuinfo.patch
  * removed-external-toolchains.patch
- Major changes are:
  * tf.distribute introduces experimental support for asynchronous training of
    models via the tf.distribute.experimental.ParameterServerStrategy API.
  * MultiWorkerMirroredStrategy is now a stable API and is no longer considered
    experimental.
  * Introduces experimental support for a new module named tf.experimental.numpy
    which is a NumPy-compatible API for writing TF programs.
  * A major refactoring of the internals of the Keras Functional API has been
    completed, that should improve the reliability, stability, and performance
    of constructing Functional models.
  * Keras mixed precision API tf.keras.mixed_precision is no longer experimental
    and allows the use of 16-bit floating point formats during training,
    improving performance by up to 3x on GPUs and 60% on TPUs.
  * TensorFlow Profiler now supports profiling MultiWorkerMirroredStrategy and
    tracing multiple workers using the sampling mode API.

- Lower memoryperjob to 1300 MB (as done for tensorflow)

- fixed hpc flavor and Leap15.2 builds

- updated to 2.1.2 with following fixes (bsc#1177022):
  * Fixes an undefined behavior causing a segfault in tf.raw_ops.Switch (CVE-2020-15190)
  * Fixes three vulnerabilities in conversion to DLPack format
    (CVE-2020-15191, CVE-2020-15192, CVE-2020-15193)
  * Fixes two vulnerabilities in SparseFillEmptyRowsGrad
    (CVE-2020-15194, CVE-2020-15195)
  * Fixes an integer truncation vulnerability in code using the
    work sharder API (CVE-2020-15202)
  * Fixes a format string vulnerability in tf.strings.as_string (CVE-2020-15203)
  * Fixes segfault raised by calling session-only ops in eager mode (CVE-2020-15204)
  * Fixes data leak and potential ASLR violation from tf.raw_ops.StringNGrams (CVE-2020-15205)
  * Fixes segfaults caused by incomplete SavedModel validation (CVE-2020-15206)
  * Fixes a data corruption due to a bug in negative indexing support in TFLite (CVE-2020-15207)
  * Fixes a data corruption due to dimension mismatch in TFLite (CVE-2020-15208)
  * Fixes several vulnerabilities in TFLite saved model format
    (CVE-2020-15209, CVE-2020-15210, CVE-2020-15211)
- using fft2d.tgz instead of fft.tar.gz
- removed fft.tar.gz

- fixed json-feature-name.patch for leap15.2 builds

- updated disk constraints, as sometimes the build fails with too
  low disk space

- Package C-headers for standard tensorflow (boo#1175789)
- fixed build gcc10.1 errors for Tumbleweed with following
  upstream patch:
  * added file Provide-overload-to-cope-with-const-ness-change-of-N.patch

- Package header files for Tensoflow2 Lite - boo#1175099

- Revert memoryperjob constraint support and use again
  %limit_build macro to avoid OOM errors

- Lower memoryperjob to 1300 MB (as done for tensorflow)

- Use memoryperjob constraint instead of %limit_build macro.

- fixed build with json_cpp 1.9.3 (bsc#1173314)

- fixed local CUDA builds

- updated to 2.1.1 which is a bug fix release mostly for external
  sources which are not part of this package (sqlite,libjpeg-turbo,
  Apache Spark)
  * Fixes a versioning bug which causes Keras layers from TF 1.x to
  be used instead of those from TF 2.x

- fixed broken builds which were caused due to missing dependency
  on @com_google_absl//absl/strings in various BUILD files
- added patch: fix-google-absl-memory.patch

- added mkl-ddn as source and do not use system mkl-dnn (bsc#1168839)
- removed patches:
  * fixed-mkl-sgemm-call.patch
  * added-mkl_dnn-as-syslib.patch
- added source:
  mkl-v0.21.2.tar.gz

- tensorflow2-lite-devel does not requires libtensorflow*

- removed hpc-mvapich2 build (bsc#1167735)

- Use pip install --no-compile (boo#1094323)

- Lite flavor should not provide python3-tensorflow nor tensorflow

- removed sources of bazel sources and replaced them by internal packages
  * rules-cc.zip removed
  * bazel-toolchains.tar.gz removed
  * bazel-skylib.0.8.0.tar.gz removed

- Lite flavor should not provide "tensorflow", otherwise
  tensorlfow2-devel and tensorlfow2-lite-devel conflict and break
  armnn

- added Provides: tensorflow, so that Kerase works with this package
  and fixed Leap 15.2 build

- Fix name for libtensorflow* sub-packages

- openSUSE has no CUDA package, so disable cuda build for
  openSUSE

- addding changes for CUDA builds

- Add 'Provides' only for hpc flavors, otherwise it matches the
  package name

- Add provides/conflicts to avoid to install tensorflow and
  tensorflow2 as some files are provided by both packages

-  removed mkl-dnn as sourc and force usage of system mkl-dnn
  for x86_64 builds
  * removed file mkl-dnn-v021.2.tar.gz
  * added patch: added-mkl_dnn-as-syslib.patch
  * added patch: fixed-mkl-sgemm-call.patch

- Add 1.25.0 as minimal version for grpc-devel

- Add 3.1.5 as a minimal version for double-conversion-devel
  (2.0.1 from SLE15SP2/Leap15.2 is too old)
- Lower required version for protobuf (3.9.1 from SLE15SP2/Leap15.2
  is fine)

- removed AVX2 flavor, this should be fixed via mkl-dnn

- Fix build on hpc targets

- added shared library packages libtensorflow2,
  libtensorflow_cc2 and libtensorflow_framework2
- removed the AWS sdk support as this forces a SEGFAULT
  * remobed file aws-sdk-cpp-1.5.8.tar.gz
- dropped following source files as they are not needed any more
  * removed file backports.weakref-1.0rc1.tar.gz
  * removed file gettid.patch
  * removed file grpc-v1.24.2.gz
  * removed file libjpeg-turbo-2.0.0.tar.gz
  * removed file nsync_1.20.0.tar.gz