* test version of matmul_v2

* added matmul_v2 grad kernel

* minor changes

* minor changes

* minor change for CI approval

* CI fix

* CI fix

* trigger CI

* changes after review, not working yet

* moved ops to anonymous namespaces

* changes after review

* test version of matmul_v2

* added matmul_v2 grad kernel

* minor changes

* minor changes

* minor change for CI approval

* CI fix

* CI fix

* added squeeze and squeeze2 kernels

* CI fix

* CI fix

* CI fix

* disabled tests when compiled with cuda

* added setting format_tag by strides

* added sigmoid BF16 FWD/BWD and gelu BF16 BWD

* changes after review

* Revert "added sigmoid BF16 FWD/BWD and gelu BF16 BWD"

This reverts commit 6e3f76720b545abfcff9f6052b46b73a1e745cae.

* Revert "Merge branch 'matmul_v2_grad' into squeeze2_op"

This reverts commit 06fcf67843a4a7884eccdf67a02a03575e1d4cb8, reversing
changes made to 6e3f76720b545abfcff9f6052b46b73a1e745cae.

* minor change

* added reshape1/2 kernels

* moved some functions into private block

* CI fix

* CI fix

* CI fix

* added expand_v2 bf16/fp32 kernel

* minor change

* CI fix

* added missing test file

* added formatting

* reduced binary size

* CI fix

* added sigmoid BF16 FWD/BWD and gelu BF16 BWD

* added newline at EOF

* switched from lambdas to local functions

* changed function names

* added prelu bf16/fp32 fwd/bwd kernel

* added matmul_v2 bf16/fp32 FWD kernel

added matmul_v2 bf16/fp32 FWD kernel

* added formatting

* removed some tests due to timeout in CI

* refactored tests

* merged tests classes into one file

* minor change

* removed test guard for CUDA

* remove skipIf

* changes after review

* formated one file

* minor change

* added skipping UT in CUDA place

* - fix to #33282

* - Increased threshold for elementwise_mul_bf16 grad

* -disabled faulty UT

* - fix to approval

* base changes for split op

* 90% of split functionality added

* full fp32 functionality

* added bf16 test

* added submemory caching

* added bf test to static mode whitelist

* minor change

* enabled split op for inference

* minor fix

* minor fix

* Add oneDNN AXPY handler.

* Add fallback for small tensors.

* Fix ifdefs

* Remove unnecessary namespace prefixes and add missing headers.

* Guard handler_axpy with proper ifdefs.

* Compilation of this function is possible only when Paddle is not build
with CUDA nor HIP.

* Move AXPY handler code to separate files.

* Use oneDNN AXPY handler in SGD op.

* Use axpy handler only when Paddle is built with oneDNN.

* Add test for SUM BF16 with big rows.

* Fix SFINAE rules for elementwise_add_to.

* Add test case for SGD with big rows.

* update

* update
Co-authored-by: NAdam Osewski <adam.osewski@intel.com>

* - Draft of implementation of refactoring

- compilation fix

* - Fixes after review

* - Removed unnecessary comment

* added op cast functionality for fp32/bf16

* added newline

* added entries in static mode white list and unity build

* fixed failing tests

* changes after review

* added formatting

* upgraded tests file as reviewer suggested

* changes after review

* minor change

* added support for most matmul cases

* added more functionality

* full functionality of matmul op, fp32 only

* added bf16 tests and functionality

* added formatting

* changes after review

* minor change

* added reviewers suggestions

* - Added clearing oneDNN per executor

* - Executor is nt always having FLAGS_use_mkldnn set to true

* fix cache key in concat oneDNN kernel

* key simplified

* - introduced fix onednn 2.2 version

* - compilation fix

* Unification of bfloat16 enablement process and refactor

* Remove unnecessary function

* Standardize the output name search

* Add conv transpose BF16

* Share function GetWeightsTz

* Adjust to review and fix op compatibility

* Add bias to unique handler name

* Remove errors related to paddle enforce

* Add conv2d_transpose to bf16 list and kernel refator

* modify error message based on comments

* edit code according to review.

* Correct spelling according to review.

* - Added UT for testing elementwise_mul caching

* lint fixes

* Add mkldnn nearest_interp and bilinear_interp op
* don't run mkldnn interpolate in default
* add interpolate_mkldnn_pass