* Fix quantize model deploy bugs when using MKLDNN (#45920)

* fix immutable op quantize bugs

* fix

* fix build bug

* fix test

* notest,test=inference

* fix ppyoloe acc drop bugs

* fix test

* fix test

* add test

* fix

* fix

* fix test

* fix refined name bug

* fix test

* bias fix

* fix matmul weight dequant bug

* re-ci

* fix tester

* fix test

* fix tester

* update weight dequantize func

* update code

* update test for converage

* update test

* update cmake

* update cmakelist

* update code

* rerun ci

* remove useless code

* re-ci

* update code

* update code

* fix header

* update code for log

* Enhance the layernorm shift partation fuse op when shift size > 0 (roll shifting)
* fix cherry-pick test

* first commit

* conver done

* correct format

* layernorm_shift_partition

* correct convert

* redefine plugin

* runable

* bug fix

* modify ShiftPartitionPattern

* correct

* add UT

* modify ut

* compile

* modify enforce

* modify UT

* Add int8 support for matmul+elementwiae_add fuse

* Corrections after review and ernie test fix

* merge conv_concat_relu to conv_act

* fix typo

* extend unit test

* reuse existing gpd

* codestyle

* enforce mkldnn conv

* convert multihead to oss

* fix:bug

* fix:delete const cast

* fix:don't support bias_qk

* add vit pass

* fix:convert bug and add preln_residual_bias

* support length=-1

* add UT for convert

* add no_bias_qk support for gpu_multihead_op

* delete infer_shape depends on bias_qk

* oss just can be used in T4 and A*

* fix:change api for ROCM CI

* Add unit tests

* matmul_v2 + activation

* matmuls + elementwise_add

* matmul_v2 postops

* transform matmul to v2

* opcompat

* fix fusing matmul with multipe outs

* add shape constraints

* remove unused vars

* change pass order

* - Unit tests to be debugged

- fix

- refactor

- diagnostic

- more diagnostic

- fix

- Fix number two

- fix

- fix

- fix

- alpha added

- more fixes

- compilation fix

- removed diagnostic code

- cosmetic fixes

* lint

* add alpha constraint

* merge matmul refactor

* trigger CI

* - fix

* - another fix

* code style

* add support for matmul+elementwise_add+activation

* code style

* fix bfloat16 bugs

* change append_binary to append_sum
Co-authored-by: NJacek Czaja <jacek.czaja@intel.com>

* fix RemoveNode in fuse_elewise_add_act_pass

* fix

* change pointer to share_ptr

* fix

* fix

* fix format

* fix

* fix graph_safe_remove_nodes

* Rename BOOST_GET macros

* Fix conflicts

* Quantize shape operator

* Add shape op to propagate scales pass

* reduce redundancy

* python code style

* fix int8 ut

* Refactor quantization of immutable ops

* Fix code formatting

* Fix formatting

* Specify input names

* Fix formatting

* Change string to reference

* Formatting

* modify graph_pattern to thread_local

* modify graph_pattern to thread_local

* sylwek

prototype params to int8 pass

* trying to make warmup work

* wip

* wip

* change test to cpp test

* review fixes, refactoring

* more refactoring

* add erasevars

* change test to fixture

* rename pass

and reorder erasevars and graphsaferemovenodes

* fix

* more refactoring and fixed bug

* formatting

* remove scale count

* enfroce message too short

* remove erasevars

erasevars couldbe cauuse of memory issues

some other fixes

* add count of successfull fuses to name of new nodes

* FindVar -> GetVar and use ConvResidual pattern

* use tensor->clear() instead of new variable

* Update paddle/fluid/framework/ir/mkldnn/params_quantization_mkldnn_pass_tester.cc
Co-authored-by: NTomasz Socha <tomasz.socha@intel.com>

* Update paddle/fluid/framework/ir/mkldnn/params_quantization_mkldnn_pass_tester.cc
Co-authored-by: NTomasz Socha <tomasz.socha@intel.com>

* Update paddle/fluid/inference/tests/api/analyzer_lexical_analysis_gru_tester.cc
Co-authored-by: NTomasz Socha <tomasz.socha@intel.com>

* add log (review fix)c

* review fix (2 functions to one)

* code review: Conv->QuantizeConv

* revert

* fix formatting

* remove unused functions

* add paddle enforce
Co-authored-by: NTomasz Socha <tomasz.socha@intel.com>

* Add elementwise_sub quantization

* Remove unnecessary comments

* Specify names for tests

* Remove comments

* Remove comments leftovers

* support yolov5s static/int8

* fix eltwise_sub and div weight compute

* fix delete_fill_constant_pass

* Add int8 scales gathering pass for convolution

* Fix typo

* Add unittest

* Add corrected unit test

* Change test name

* Remove enabling mkldnn in test

* Speed up test

* Change max examples

* Add functional test

* Change test name

* Add new test case

* Rename pass

* pdnode_compare

* panode compare

* pdnode_compare

* Refactor Quantization

* Refactor Dequantization

* Classy solution

* Style I

* Style II

* Style III

* Use VLOG(4) for debug info

* Style IV

* DRY

* change nodes names

* add const prefix

* change asX to as_x in all files

* paddle inference support new quant_model

* Quantize elementwise mul op

* Parametrize elementwise functions

* Fix code formatting

* Add elementwise add and activation fuse pass

* Fix copy ellision

* More flexible pattern detector

* More flexible fusion pass

* Update lists for pass

* Add support for Pow operator

* Add support for more activation types

* Style

* Rename fusion pass

* First version of tests

* Dirty version of pass

* Polished version

* Update pbtxt

* Style

* Update names

* Style

* Use PADDLE_ENFORCE_EQ

* Save error message to variable

* WO for error checks

* CR

* Static style check

* Add missing 'activation_scale' attribute

* Add relu6 and sigmoid activations

* Style

* Fix fuse list formating

* Sync filenames for fuse pass files

* Fix cmake after move

* Fix registration

* Fix pass name in tests

* Add missing activations to checker

* WIPS

* Working mul op

* Working sub

* Working Add

* Remove pten includes

* Remove some forward declarations

* Remove Includes

* Fixes

* Remove default kernels

* Add check if post_ops attributes are avaliable

* Style

* Code adjustment

* Register default kernels

* We have year 2022 not 2021...
Co-authored-by: Njakpiase <jakpia21@gmail.com>
Co-authored-by: NSylwester Fraczek <sylwester.fraczek@intel.com>

* Fast review fixes
Co-authored-by: Njakpiase <jakpia21@gmail.com>
Co-authored-by: NSylwester Fraczek <sylwester.fraczek@intel.com>

* Review Fix

* Rename one_dnn -> onednn

* Style after review

* Fast and dirty fix for quantization

* Update tests

* Style

* Fix mkldnn_quantizer config

* Add Joanna's suggestion.

* Check if operator is explicitly disables on OneDNN

* Try to use unregistered attributes

* Style

* Test new framework

* FXI

* FXII

* Update test

* Style
Co-authored-by: Njakpiase <jakpia21@gmail.com>
Co-authored-by: NSylwester Fraczek <sylwester.fraczek@intel.com>

* Added cuBlasLtHandle_t to device context.

* Added fused_gemm_epilogue op.

1. Added fused_gemm_epilogue op to leverage cuBlastLt Epilogue.
2. Support fusion Act(X*Y + bias), X'dims >=2 and Y'dims shoule be 2.
2. Act currently only be supported ReLU. (Will add GeLU in the future).

* Added UT to fused_gemm_epilogue op.

* Added LinearAct Pattern

1. Added LinearAct into graph_pattern_detector.* to define (2.)'s
pattern.
2. LinearAct is used to detect act(element_add(matmul_v2(x, w), bias)).
3. act currently only support ReLU (Will support GeLU in the future).

* Added FuseGemmEpiloguePass

1, Added FuseGemmEpiloguePass to handle nn.Linear + Act{ReLU}
fusion (GeLU will be supported in the future).
2. Only support matmul_v2 from nn.Linear.

* Added pybind to BuildStrageter.fuse_gemm_epilogue_.

* Added UT for fuse_gemm_epilogue_pass.

* GeLU support and EpilogueSingleton

1. Added GeLU support to fused_gemm_epilogue op.
2. Added EpilogueSingleton to cache auxiliary pointer.
3. Added related UTs.

* Rename cublaslt_epilogue_opto gemm_epilogue_op.*.

* Added both train and infer pattern to LinearAct.

1. Added support of fwd graph with grap_ops linking to LinearAct.
2. Added related changes to fuse_gemm_epilogue_pass for above
modification.

* Changed CUDA requirement from 11.4 to 11.6 for fuse_gemm_epilogue_pass.

* Added identity activation support to gemm_epilogue_op.

* Added Linear Fusion (matmul_v2 + ele_add)

1. Added matmul_v2 + ele_add pattern to LinearActPattern.
2. Added matmul_v2 + ele_add support to fuse_gemm_epilogue_pass.

* Rename gemm_epilogue_op.* to fused_gemm_epilogue_op.*

* Add fused_gemm_epilogue_grad op.

1. Added fused_gemm_epilogue_grad to support backward epilogue fusion.

* Add UTs to fused_gemm_epilogue_grad_op.

* Change attribute name in fused_gemm_epilogue_grad_op for clearing.

* Allow DX and DBias be dispensable to fused_gemm_epilogue_grad op.

* Added ElementwiseAdd+Matmul+Act graph pattern detection.

* Fuse backward of Linear( Act(x))

1. Added backward fusion pass to Linear( Act(x)).
2. Added backward fusion pass to Linear(x).

* Added UTs to backward fusion of Linear(Act(x)).

* Complete document of arguments to fused_gemm_epilogue_op.

* Made arguments of some functions pass by reference.

* Modify code with review comments.

1. Made arguments of some function pass by reference.
2. Removed redundant code.
3. Followed Google code style to change code.

* Made 'const' code style be consistent

* Fixed random seed of python UTs.

* Set Compiling constrains to cuBlasLt

1. Require CUDA 11.6+
2. Remove fuse_gemm_epilogue related tests when CUDA < 11.6.

* Code Reivew from Paddle

1. Changed arguments name is_first_gemm to without_x_gradient for
clearing.
2. Applied PADDLE_THROW in fused_gemm_epilogue_op.

* Remove EpilogueSingleton

1. Applied ReserveSpace to replace Epilogue for passing auxiliary
pointers between FWD and BWD.

* Fix a logical error and enhance UTs.

1. Added act op count checking in UTs.
2. Fix issue to fuse backward or ReLU(Linear(X)).
3. TODO: solve GELU fusion issues.

* Fix Linear and GeLU fusion issues.

1. Modified graph_detech_pattern to fit with both linear wiht gelu or
relu.
2. Modified data range in Uts to allow negative values.

* Removed fused_gemm_epilogue_op.h.

* Rename namespace pten to phi.

* Rename name of arguments in fused_gemm_epilogue_op

1. bias -> Bias.
2. out -> Out.
3. reserve_space -> ReserveSpace.

* Change EpiloguePassActivationCache as local variable.

1. Removed singleton in EpiloguePassActivationCache.
2. Made EpiloguePassActivationCache as an argument to each pass
functions.

* Fix for split bf16 inference

* added test for pass

* changes after review

* Fix quantization next op findings

* Corrections according to the review

* Quantize nearest_interp and nearest_interp_v2

* Check if avx_core supported

* Add depthwise_conv2d to supported quantization list

* add matmul_scale matmul_v2_scale fuse pass

* add scaletensor judge

* modify var name

* add timeout notest;test=coverag

* fix error commit

* fix use_mkldnn attr

* fix use_mkldnn attr

* remove bf16

* remove comments

* remove wrong return

* fix UT

* reshape+transpose+matmul_v2

* in_name->input_name

* fix pr-ci-static-check

* quantize slice op

* correct test

* fix code formatting

* added softplus + activation fuse plass

* minor change

* implemented reviewer suggestion

* minor fix

* minor fix

* added scale_out parameter

* minor fix

* fix for iScan CI

* conditionally disabled logs

* refactored pass builder

[Paddle-Inference]Add MatmulV2ToMatmul convert Pass, fix (matmul_v2, matmul, mul) convert pass, fix (matmul, mul) op_teller (#36652)

* new_Matmul2ToMatmulToMul

* new_Matmul2ToMatmulToMul

* fix paddle_pass_builder

* fix paddle_pass_builder

* fix paddle_pass_builder

* tem

* tem

* Add MatmulV2ToMatmul convert Pass; MatmulV2ToMul convert Pass

* Add MatmulV2ToMatmul convert Pass; MatmulV2ToMul convert Pass

* add matmul_broadcast_unitest

* fix op_teller

* added base changes for matmul_v2+trans+resh fuse pass

* added full matmul_v2+transpose+reshape pass

* removed a file added by mistake

* added reviewers suggestions

* Changed ops type in checking capatibility version

* Deteled one statement

* support bert when exists matmul_v2

* update