GitLab

add flags for ck, Enable CK with hipRTC.  CK can be used with the MIGRAPHX_ENABLE_CK=1 and MIGRAPHX_TUNE_CK=1

Move CI to support the rocm5.5 release

* Add hiprtc driver

* deprecate HCC

Added CMakeFlag for hipRTC. MIGRAPHX_USE_HIPRTC.
Added stages in Jenkins for hipRTC.
Fixes for some of the pending issues from hipRTC.

HIP had change in previous rocm releases to use --offload-arch instead of --cuda-gpu-arch.

This should be backwards compatbile. hipRTC also supports --offload-arch.

Improve the assembly dump to track where certain instruction come from.

Updated each source file in the repo with the existing license.

* Upgrade to cppcheck 2.8

Improves performance for add_gelu.  In bert it is 4x faster and for mul_add it is 50% faster than what we current have.

Fixes the #error when using cppcheck. This no longer suppresses cppcheck errors when including those errors. This fixes the cppcheck errors that was there already.

There is significant improvement on larger tensors with half almost 50% faster:

lens: [1024, 384, 768]
gpu::code_object[code_object=13832,symbol_name=kernel,global=39321600,local=256,]: 1.16685ms
gpu::reduce_sum[axes={2}]: 1.73126ms
Also for non-trivial layouts this can sometimes be over 2x faster:

lens: [64, 1024, 768, 4]
gpu::code_object[code_object=13832,symbol_name=kernel,global=39321600,local=256,]: 1.1706ms
gpu::reduce_sum[axes={1}]: 2.63375ms
Of course if the stride becomes larger this speed improvement diminishes due to poor memory access patterns. A lane_reduce instead of a block_reduce is needed for such type of kernels. I plan to address that in a future PR.

Finally, this also includes a MIGRAPHX_GPU_DUMP_ASM env variable which will print out the assembly when the kernel compiles.

This adds the infrastructure so we can compile everything in parallel, whereas before only pointwise kernels were compiled in parallel. This will also directly integrate with lowering and the gpu-driver. The kernels for pointwise and roialign are using this infrastructure. Scatternd is not since it does require standard shape.

This also makes it easier to add new runtime compiled kernels in the future.

* Use ccache for runtime compilation

Boost the max number of workgroups for pointwise ops by matching what we are doing in launch.hpp

GPU implementation of the roialign operator, using the jit approach to reduce the lib size.

* Add hiprtc compile option
* Add cross compile test
* Update error reporting
* Add tests for errors and warnings
* Fix tidy warning
* Add comment to ifdefs
* Skip null character at end of log
* Assert there is null at the end

* Add method to compile pointwise

* Formatting

* Add lambda

* Add semicolon

* Rename variable

* Add driver to run jit kernels

* Formatting

* Add context

* Formatting

* Make seperate driver folder

* Add more general gpu driver

* Formatting

* Print out wll time

* Formatting

* Run multiple times and skip first run

* Formatting

* Seperate time_op

* Run an op for comparison

* Formatting

* Add debug asserts

* Formatting

* Change parameer name

* Formatting

* Fix argument order

* Formatting

* Add preloading

* Formatting

* Allow a different data type

* Formatting

* Pipeline transformations

* Formatting

* Add vectorization

* Formatting

* Reduce dims

* Formatting

* Compile with launch params as constant

* Formatting

* Make sure buffer can be vecotrized

* Formatting

* Enable vectorization and preloading

* Formatting

* Add print header

* Formatting

* Avoid allocating to large of LDS

* Formatting

* Add some vec functions to a seperate header

* Formatting

* Add stride loops

* Formatting

* Improve the transform pipeline

* Formatting

* Add const

* Fix shape check

* Formatting

* Just check stride axis is zero

* Remove extra finc_vector_axis overload

* Simplify some mroe functions

* Formatting

* Remove some more extra functions

* Formatting

* Simplify more decltypes

* Add another const

* Fix test

* Get buffer pointer different for older compilers
Co-authored-by: Shucai Xiao <shucai@gmail.com>
Co-authored-by: Chris Austen <causten@users.noreply.github.com>

* Add definitions for all pointwise operators

* Formatting

* Add cpp generator class

* Formatting

* Move compilation to core

* Formatting

* Add clock to tmp name

* Add dynamic loader

* Formatting

* Add tests for code gen

* Formatting

* Add test for literals

* Formatting

* Use with_char

* Add missing header

* Fix mismerge

* Ignore tidy warning

* Fxx gcc 5 errors

* Apply fixits

* Skip signed bitwise of status

* Remove unused parameters

* Explicitly add c++14 flag

* Fix tidy warning

* Remove .o files

* Add code object op

* Formattting

* Add more value tests

* Formatting

* Fix from_value conversion from binary

* Formatting

* Dont use offload copy

* Remove iostream header

* Fix compilation errors

* Formatting

* Rename var

* Add missing files

* Formatting

* Remove duplicate variable

* Remove comment

* Template the function so sfinae will work

* Formatting

* Use template specialization since ADL is broken on hcc

* Formatting

* Annotate the constructor with HD for hcc

* Make variable const
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>

* Add compiler flags

* Add missing include

* Add filesystem header

* Formatting

* Add tmp_dir to run

* Formatting

* Kernel compilation and launching

* Formatting

* Seperate pack_args

* Formatting

* Add alignment tests

* Formatting

* Add compile test

* Formatting

* Complete compile test

* Formatting

* Use is_regular_file free function

* Fix is_regular_file call

* Fix tidy issues

* Fix tidy

* Fix tidy issue

* Print size in read_buffer to debug issue on jenkins

* Add hip flags before src file

* Fix reading output files

* Fix unsued variable warning

* Formatting

* Formatting

* Disable tidy check
Co-authored-by: Shucai Xiao <shucai.xiao@amd.com>
Co-authored-by: mvermeulen <5479696+mvermeulen@users.noreply.github.com>