GitLab

* Add basic support for direct loads from global to LDS

* Clean the code and comments

* Add support for fp16

* Add comments

* Add check for thread cluster lengths

* Align non-direct-load fp16 example

* Small fixes

* Extend IsSupported to check for supported GPU gens

* Build examples only on the supported HW

* Do not throw when instance not supported in 04 example

* Review: Apply review suggestions

* Review: small fix

* Review: small fix

* improve kpad

* more tuning parameters

* f16_f8_fp16

* cut test time

* add f16_f8_fp16

* add f16_f8_f16

* testing instances for skinny cases

* format

* clean

* add fp16_f8_fp16

* clang-format

* add grouped gemm instalces

* fixed profile grouped_gemm

* clean

* clean

* clean

* clean

* clean

* add missing instance func

* fixed inferface

---------
Co-authored-by: Jing Zhang <jizha@amd.com>
Co-authored-by: root <root@sh5-1e707-rc06-38.mkm.dcgpu>

* Fix the fp8 conversion

* Try clipping value before conversion

* Fix return

* Simplify with a const

* reduce the gemm input tensor values to reduce round-off error

* replace if-else with lambda

* fix syntax

---------
Co-authored-by: Rostyslav Geyyer <rosty.geyyer@amd.com>

* Fix the conversion

* Add bf8 functionality

* Enable example on MI200 as well

* apply the patch for dl kernels on gfx11

* build DL kernels on navi32 CI

* Extend available elementwise operations with conv examples

* Fixes

* Remove not needed convert

* Update CMakeFile and dir name

* Add a condition to build fp8 instances

* simplified buffer_load/store

* add bfp8/fp8

* fixed

* remove all f8/bf8 condition include folder

* fixed cmake conditions

* fixed DTYPES=fp16/bfp16

* fix

* fixed buffer_load

* fixed buffer_store

* fix

* clean example cmake files

* fixed ci

* fixed cit

---------
Co-authored-by: Rostyslav Geyyer <rosty.geyyer@amd.com>
Co-authored-by: Jing Zhang <jizha@amd.com>

Co-authored-by: Jing Zhang <jizha@amd.com>

* add vector_type support into thread_copy_v3r1

* remove unncessary type_convert

* fixed datatype

* fixed dataType

* changed API with is_packx_invocable

* changed example

* add missing cmake file

* fixed ci

* fixed cmake

---------
Co-authored-by: Jing Zhang <jizha@amd.com>

* simplified buffer_load/store

* add bfp8/fp8

* fixed

* fixed buffer_load

* fixed buffer_store

---------
Co-authored-by: Jing Zhang <jizha@amd.com>

This reverts commit a4f72a31.

* Introduce LocalBlockToCTileMap.

* Change the signature of CalculateBottomIndex() function which now does
not accept any argument. The B2C map which is already passed as an
argument to the kernel Run function is calculating block's local id
already outside at kernel entry point __global__ function.
The LocalB2C map stores as members local block ID.

* Use LocalBlockToCTile map in device ops.

* First draft of tile loop work distribution.

* Fix typo.

* Simplify kernel arguments.

Calculate descriptors & B2C maps on the device.

* Use looping kernel.

* Fix B2C constructor.

* Fix Navi21 errors.

* Calculate tile start/end in device kernel.

* Change Run API to accept user provided workspace buffer.

* Add new line at EOF.

* Move Gemm KernelArguments to device op interface.

* Remove unused code.

* Update API.

* Launch grid size which is min of occupancy vs tile count

* Get back to use constant memory for gemm descriptors.

* Remove unused code.

* Add default virtual method implementation.

* Update comments to conform with doxygen style.

* Fix doc style and unused parameters.

* Add thread cluster lengths to kernel name.

* Remove old splitk impl and replace it with tile looping one.

* Modify instances.

* set KPerBlock to 64
* maximize wherever possible vector load size.

* Fix instances cluster lengths.

* Change comment style.

* Use 128b store where possible in instances.

* Update test cases, since KPerBlock has doubled.

* Update output stream operator for Sequence.

* Add pipeline version to GroupedGEMM device op type string.

* Fix pipeline version type logging.

* Fix input tensors type after merge.

* Fix compiler error.

* Fix output stream operator for Pipeline version.

* Store using 128b.

* Set of instances with kpb 32/64

* Limit number of instances

* Remove commented out instances.

* Fix function name.

* Limit the number of instances.

Add pipline version to the regular instances

* Change thr cluster layout for reading B tensor.

* disabled failed instances

---------
Co-authored-by: Adam Osewski <aosewski@amd.com>
Co-authored-by: zjing14 <zhangjing14@gmail.com>
Co-authored-by: Jing Zhang <jizha@amd.com>

* workaround nan problem by changing output to fp16

* enable f8/bf8 gemm tests on MI200

* workaround f16 to f8 conversion

---------
Co-authored-by: Jing Zhang <jizha@amd.com>

* Add f8 bf8 gemm example

* Add element-wise ops

* Add intrinsics

* Update reference calculation

* Add an additional type option for xdlops gemm

* Fix build process

* Add bf8 to buffer addressing

* Update blockwise op, split typeA and typeB

* Update for compatibility

* Uppdate naming to f8->fp8

* Update naming

* Format

* Update naming (#937)

* Add a client example

* Add computetypes to device and gridwise ops

* Add instances, update instance factory

* Format

* Fix a flag

* Add ckProfiler mode

* Fix typos

* Add an example

* Add bf8 generator

* add bf8 mfma; fixed type_convert for bf8

* move verfication ahead of timing

* Update reference calculation

* Fix reference

* Narrow down float init range

* Fix bf8 bf8 mfma

* Add bf8 @ fp8 mfma

* Update example

* Update instances

* Update profiler api

* Update for compatibility

* Format

* Remove extra example

* Clean up

* workaround convert

---------
Co-authored-by: Jing Zhang <jizha@amd.com>

* Add f8 bf8 gemm example

* Add element-wise ops

* Add intrinsics

* Update reference calculation

* Add an additional type option for xdlops gemm

* Fix build process

* Add bf8 to buffer addressing

* Update blockwise op, split typeA and typeB

* Update for compatibility

* Uppdate naming to f8->fp8

* Update naming

* Format

* Handle type conversions to a const datatype

* Review: Handle X being const data type as well

* Review: Remove typo

* Add column to image kernel

* Minor fixes for dtypes and client examples

* Disable tests for disabled dtypes

* Disable add instances functions for disabled data types

* Minor stylistic fixes

* Revert "Disable add instances functions for disabled data types"

This reverts commit 728b8695.

* Instances reduction

* Add comments in device_column_to_image_impl

* Update changelog and Copyrights

* Improve changelog

* add gridwise_multi_abd

* move element_op into RunRead

* merge element_wise op with data read

* add multiABD example

* allow packed elementwise_op

* changed example

* clean

* clean

* add is_detected

* fix

* minor fix

* add scaleAdd_vec4 example

---------
Co-authored-by: Jing Zhang <jizha@amd.com>

* Add native conversions

* Add bf8 conversions

* Add grouped conv bwd weight dl instances and new layout

* Add M and N padding

* Remove todo comment

* Enable grouped conv fwd dl k,c=1 generic instance

* Comment fixes

* Refactor f8_t to add bf8_t

* Add check_err impl for f8_t

* Update fp8 test

* Format

* Revert the fix

* Update vector_type implementation

* Add bf8 test

* Add bf8, use BitInt types

* Add bf8 conversion methods

* Update type_convert for fp8/bf8

* Add check_err fp8/bf8 support

* Add subnorm fp8 tests

* Add subnorm bf8 tests

* Fix conversion

* Add bf8 cmake bindings

* Add macros to enable build with disabled fp8/bf8

* Remove is_native method

* Update flag combination for mixed precision instances

* Add more flag checks

* Add another flag to a client example

* Add type traits, decouple f8/bf8 casting

* Clean up

* Decouple fp8 and bf8 flags

* Remove more redundant flags

* Remove leftover comments

* Redesign the DPP8 GEMM kernel to use warp-wise component

* Review: Improve error messages

* Review: Remove unnecessary empty lines

* Review: Fix M, N per thread names

* Review: Rename mfma_input_type to dpp_input_type

* Review: Fix tensor adaptor; remove unnecessary element

* Review: Remove calls to dpp_gemm's MakeCDescriptor

* Review: Add blockwise doc, change function names to include dimension names

* Review: Remove duplicated code; Move Block2CtileMap alias to the top of the file

* Review: Add __restrict__ keywords

* Review: Use MatrixPadder for padding A, B, C matrices

* Review: Remove hardcoded datatypes

* Review: Change names from FloatX to XDataType

* Review: Introduce AK0 and BK0 instead of a single K0

* Review: Remove construction of dpp_datatypes object

* Review: Rename DppInstrRunner to DppLanegroupGemm

* Add maxpool instances

* Rename index pool to max pool.

* Add maxpool bwd bf16 instances

* Add avg pool bwd instances

* Rename avgpool and maxpool to avg_pool3d and max_pool

* Add bf16 pool fwd instances

* Add max pool bwd to ckProfiler

* Add avg pool3d bwd to ckProfiler

* Add avg pool bwd test

* Fix bug of reference pool fwd (dilation)

* Fix bug of max pool bwd  (dilation and initZero)

* Support bf16 compute data type

* Force compute type be f32. Because atomicAdd only support f32

* Add max pool bwd test

* Rename folder

* Rename pool

* Add max pool bwd client example

* Add avg pool bwd client example

* Add missing workspace

* clang format

* Rename macro

* remove useless header

* remove useless layout

* Add s_nops after v_dot to avoid hazard

* Fix builtin for inner_produxt fp16

* Skip inline version to builtin

* Add comments regarding isa

* Fix comment regarding s_nop

* initial stream-k implementation with example

* fix unexpected change in err

* improve a little bit performance by reorganize pipeline.

* improve perf a little bit by swizzle block idx

* add profiler

* update example

* fix spelling

* shrink karg for streamk

* support dynamic buffer using memory coherence glc_slc bit from template

* control memory coherence while construct dynamic buffer

* update reduction for streamk(not ready yet)

* Add template parameter to make_dynamic_buffer to support amd_buffer coherence setting

* fix build issue

* fix several bug

* now result is correct, everything works (but has scratch)

* remove scratch by manually reset coordinate

* update device code

* fix a bug in final reduce

* fix something in example

* update async memset

* fix enum as camel case

* modify coherence enum name

* clean code and use atomic streamk by default

* remove unused var

* throw exception if have empty pointer

* fix format

* fix CI warning

* fix type in init

* modify CI error

* filter out on gfx10+

* restore changed example code

---------
Co-authored-by: Qianfeng Zhang <Qianfeng.Zhang@amd.com>

* Use dim 0 as faster dim for writing mean/var/count workspace in batchnorm multiblock method [performance]

* Add CountDataType as template parameter in blockwise_welford

* Add utility/get_shift.hpp

* Add BatchNorm multiblock single-kernel implementation

* Add smem inline assembly based implementation of gms_init/gms_barrier/gms_reset for gfx90a

* Renaming in device_batchnorm_forward_impl.hpp

* Tiny fix in the batchnorm_fwd profiler

* Revert "Add smem inline assembly based implementation of gms_init/gms_barrier/gms_reset for gfx90a"

This reverts commit d16d00919c43f10759e7b4e4d112125221ed9064.

* Use the old two-kernel batchnorm multiblock method for gfx1030

* Use the old two-kernel batchnorm multiblock method for gfx908

* use the single-kernel batchnorm multiblock method only for gfx90a

* Remove get_wave_id() from utility/get_id.hpp since it is not used

* Set true for testing running mean/variance and saving mean/invvariance in the examples

* Fix to copy-right words

* Remove un-needed including in utility/get_id.hpp

* Add comments to workgroup_synchronization.hpp

* Remove un-used codes in gridwise_multiblock_batchnorm_forward.hpp

* Renaming in the kernels

* Remove un-used kernel file

* Add fp8 xdl gemm

* Add example

* Use int8 intrinsics for buffer load/store

* Format

* Update cmakelists

* Add basic fp8 definitions and prn-generator

* Format

* Add fp8<->fp32 type_convert

* Format

* Split type_convert and cast_to/from_f8

* Format

* Minor fix

* Minor fix

* Move fp8 utils to a separate header

* Add elementwise ops

* Add fp8_convert_sr

* Format

* Add element op

* Eliminate magic numbers

* Split f8_convert_sr in host and device

* Format

* Add some constexpr

* Add a datatype test

* Format

* Another format

* Add fp8<->fp16 tests

* Update type_converts

* Format

* Add fp16 casting functions

* Format

* Use seed as a runtime arg

* Use element location for PRNG

* Format

* Add fp8<->fp16 to PassThrough element op

* Clean up

* Merge host and device implementations

* Add comments on rounding modes

* Remove leftover code

* Put type_converts into a separate header

* Put random number gen to a separate header

* Rearrange f8_utils' namespaces

* Refactor type_convert.hpp

* Move f8_t definition

* enable gfx941/942 targets

* fix clang format

* fix the cmake logic for multiple targets

* fix cmake syntax for looping over targets

* add gfx941/942 support for gemm_xdl instances

* Fix wrong pointer type

* Rename type trait get_unsigned_int<> to get_carrier<>

* Add 3-bytes carrier type

* Add missing __device__ specifier

* Rename template non-type parameter

* Leave the rest byte uninitialized

* Avoid invoking (host) STL algorithms

* Remove unnecessary 'inline' specifier

* Extract common logic out as helper method

* Hide dummy member function

* Add missing __device__ specifier

* Add overloaded version of __builtin_amdgcn_readfirstlane()

* Remove 'static' specifiers

* Remove more 'static' specifier

* Replace unsigne char by std::byte

* Add 'const' specifier to never changing variable

* Add 'inline' specifier to funcion definition

* Fix wrong boundar calculation logic

* Rename type trait

* Remove std:: qualifier from standard types

* Replace 'size_t' by 'unsigned'

* Use type alias to hint usage

* Replace static_for<> by ordinary 'for' loop

* Rename readfirstlane() to amd_wave_read_first_lane()

* Rename file readfirstlance.hpp as amd_wave_read_first_lane.hpp

* Reorder statements

* fix headers for gpu instances

* remove unused headers

---------
Co-authored-by: zjing14 <zhangjing14@gmail.com>

* Add TypeConvert class and start refactoring

* Refactor TypeConvert as a struct

* Get back to template functions type_convert

* Add a type_convert_bf16_rtn, set rtz as default

* Clean up

* Add UnaryConvertPrecision struct for high-precision workloads

* Format

* Update type_convert to UnaryConvert on threadwise level

* Update UnaryConvertPrecision

* Format

* Fix chmod

* Add a flag to pick converion method

* Format

* Remove the added flag

* Merge elementwise op with type conversion

* Move type_convert to elemwise op, update the op

* Update type_convert_precision -> bf16_convert_rtn

* Clean up

* Update comments

* Update the CK_WORKAROUND_DENORM_FIX flag handling

* Update the unneeded op to work but warn user

* Remove the message

* Use a PassThrough instead of ConvertBF16RTN to calcaulate reference

* Format

* Add missing include

* enable gfx940

* switch between intrinsic mfma routines on mi100/200 and mi300

* fix mfma_int8 on MI300

* disable 2 int8 examples on MI300

* Update cmake-ck-dev.sh

* restore gitignore file

* modify Jenkinsfile to the internal repo

---------
Co-authored-by: Jing Zhang <jizha@amd.com>
Co-authored-by: zjing14 <zhangjing14@gmail.com>