[CK_TILE] Add PagedAttention kernels (#1387)

* Use dictionary to config all the functions

* Add init codegen logic for fmha fwd appendkv

* Call HIP_CHECK_ERROR() macro to get real source info

* Setup meaningfull arguments

* Sync kernel name with the codegen

* Add knew/vnew tensors to the kernel argument

* Fix wrong K values after appending

* Fix vnew append errro

* Extract common logics

* Fix Vnew tile dstr for row major case

* Conditionally add fwd_splitkv API in fmha_fwd example

* Conditionally add call to fmha_fwd_splitkv()

* Remove "EXAMPLE_" prefix of cmake variables

* Regsiter API handlers automatically

* Early return if 0 < s_k_new is not supported

* Show message if we are ignoring option

* Unify CMakeLists.txt coding style

* Set num_splits=1 if split-kv is not supported

* Add length/stride getters for HostTensor

* Add RoPE example utilities

* Add reference_rotary_position_embedding() (not implemented)

* Finish reference_rotary_position_embedding() impl

* Fix typo of HostTensor<>::get_length()

* Fix compilation errors

* Fix wrong answer when interleaved=false

* Fix wrong answer when interleaved=true

* Append K/V in the host verification code

* Simplify K appending logics

* Simplify v_host_ref definition

* Reduce input/output dimensions

* Rename function: add "batched" prefix

* Apply RoPE on host side

* Rename RoPE utility function

* Fix wrong tensor size

* Avoid invoking deprecated method 'find_module'

* Pass RoPE kernel args

* Create Rotary Cos/Sin tile windows in kernel

* Add compute data type alias for RoPE

* Randomly generate seqlen_knew if needed

* Fix seqlen_knew enabling check logic

* Add minimum seqlen_k to generate compliance kvcache

* Fix compilation error in debug mode

* Fix wrong boundaries

* Fix wrong seqlen_k for kvcache

* Rename variables used in distributio encoding

* Fix rotary cos/sin tensor/tile size

* Add constraint to the rotary_dim option

* Remove unused inner namespace

* Add dram distribution for rotary_cos/rotary_sin (interleaved)

* Only apply interleaved RoPE on Knew for now

* Fix wrong thread starting offset

* Instantiate multiple kernels for RoPE approaches

* Clean-up pipeline

* Fix error in RoPE host reference

* Handle RoPE half-rotated logics

* Support 8x rotary_dim under half-rotated RoPE

* Add comment

* Apply elementwise function to the loaded tiles

* Unify parameter/variable naming style

* Remove constness from q_ptr

* Add code blocks for q_tile

* Apply RoPE to q_tile

* Remove debug print code in kernel

* Fix wrong knew/vnew appending positions

* Use better naming for tile indices

* Add make_tile_window() for adding distribution only

* Skip code if # of block is more than needed

* Move thread locating logics into policy

* Remove always true static_assert()

* Rename header

* Rename RotaryEmbeddingEnum

* Extract rotary embedding logic out

* Re-order parameters

* Align naming of some tile size constants

* Rename more tile size constants

* Fix wrong grid size

* Fix wrong shape of knew_host/vnew_host

* Fix wrong index into knew_host/vnew_host

* Fix wrong rotary_cos/rotary_sin memory size for Q

* Extract Q/Knew vector size to helper methods

* Use different rotary_cos/rotary_sin distr for Q/Knew

* Update host/device specifiers

* Fix wrong data type for Q rotary_cos/rotary_sin

* Remove RoPEComputeDataType type alias

* Shift rotary_cos/rotary_sin by cache_seqlen_k

* Add comment for why I just 't' for all padding flags

* Align commit message to the real comment

* Fix wrong pipeline

* Rename utility function

* Disable host verification if API not exist

* Fix wrong rope key for fp8 pipeline

* Allow only apply RoPE on Q (without append KV)

* Add append-kv smoke tests

* Remove debug statements

* Remove more debug statements

* Re-arrange the 'set +x' command

* Remove no-longer used method in pipeline

* Add missing init code

* Refine pipeline padding settings

* Enlarge rotary_dim limit (8 -> 16)

* Enlarge KPerThread for rotary_interleaved=false

* Update rotary_dim range in smoke_test_fwd.sh

* Add template argument 'kIsPagedKV' for splitkv kernels

* Launch splitkv kernel if given page_block_size

* Fix wrong kernel name

* Fix seqlen_k_min for pre-fill case (1 -> 0)

* Add copy_const<> type trait

* Add another make_tile_window()

* Introduce 'TileWindowNavigator' types

* Simplify TileWindowNavigator interfaces

* Fix tile window navigation bugs

* Disable calling fmha_fwd()

* Remove ununnecessary data members

* Simplify more make_tile_window() overloads

* Move V tile through TileWindowNavigator

* Fix uneven split checking logic

* Move code after decide seqlen_q/seqlen_k

* Make sure we always start reading complete tile

* Use 128 as minimus page_block_size

* Fix wrong origin for bias

* Add batch_stride_k/batch_stride_v in group mode

* Unify origin

* Add missing kernel arguments for group mode

* Add paged-kv codegen logic for appendkv kernels

* Add block_table kernel args for appendkv kernel

* Add tile navigators to the appendkv kernel

* Fix wrong tensor descriptor lengths

* Pass re-created tile window to pipeline

* Fix wrong strides for appendkv kernel

* Allow transit tile_window to another page-block

* Handle cross-page-block write

* Donot perform write again if already in last page-block

* Always add fmha_fwd() api

* Add missing group mode argument

* Remove debug macro usages

* Rename option s_k_new to s_knew

* Separate splitkv/non-splitkv args/traits

* Remove fmha_fwd_dispatch()

* Fix compilation errors

* Remove dropout code in splitkv kernel

* Allow problem types without define kHasDropout attr

* Use generic lambda to init traits objects

* Separate more non-splitkv & splitkv traits/args

* Display more info for specific kernels

* Show more detailed warning message

* Rename 'max_num_blocks' to 'max_num_page_blocks'

* Remove no-longer used pipeline files

* Wrap code by #if directives

* Move functors to the begining of validation code

* Use generic lambda to init all the api traits/args

* Fix wrong seqlen for kvcache

* Add missing comment

* Rename TileWindowNavigator to PageBlockNavigator

* Only expose necessary methods (not attributes)

* Re-order pipeline paremeters

* Refine smoke_test_fwd.sh

* Fix wrong arugment count

* Make tile window directly via PageBlockNavigator

* Remove unused template paremeter

* Remove group mode from appendkv kernel

* Fix skcheck logic

* Fix wrong syntax in skcheck expr

* Use meaningful options in smoke test

* Remove options

* Fix formatting

* Fix more format

* Re-organize bash functions

* Pass cache_batch_idx to kernels

* Support cache_batch_idx in example

* Fix compilation error

* Add more appendkv test

* Add more case for appendkv

* Fix unexisted attribute

* Remove 0 < seqlen_knew constraint

* Clarify the case in warning message

* Remove macro checking

* Force batch mode when invoking appendkv & splitkv apis

* Fix mode overriding logics

* Fix wrong parameter name

* Randomize seqlen_k if use kvcache

* Use randomized seqlen_k for kvcache

* Avoid using too small rotary_cos & rotary_sin

* Rename parameter

* Add seqlen_q & seqlen_k rules

* Add comment

* Add more comments

* Fix compilation errors

* Fix typo in comment

* Remove type argument

* Avoid seqlen_k=0 for kvcache

* Revert "Avoid seqlen_k=0 for kvcache"

This reverts commit 21c4df89e4.

* Fix wrong uneven split checking logics

* Only randomize kvcache seqlen_k if 1 < batch

* Return earlier if split is empty

* Revert "Only randomize kvcache seqlen_k if 1 < batch"

This reverts commit b9a4ab0d7e.

* Re-order seqlen_k_start adjustment logics

* Fix compilation errors

* Re-format script

* Find executable from folder automatically

* Fix kvcache seqlen_k generating logic

* Make comment more clear

* Fix wrong knew/vew appending logic on host

* Add s_barrier to sync threads

* Revert "Add s_barrier to sync threads"

This reverts commit d3f550f30c.

* Support only using 1 row of rotary_cos/rotary_sin

* Rotate Q in different way

* Unify tensor view creation logics

* Fix wrong argument

* Add mask to switch how we use the rotary_cos/sin

* Move attr from traits to problem

* Move has_mask to fmha_fwd_appendkv_args

* Support use uint32_t as SAD operand in Alibi<>

* Use sad_u32() in splitkv kernels

* Store tensor views in PageBlockNavigator

* Use stored tensor view to update tile windows

* Enlarge tensor view size

* Remove debug code

* Fix wrong tensor view size

* Wrap tensor view into PageBlockNavigator

* Add DataType member to PageBlockNavigator

* Remove unnecessary member functions

* Refind macro use

* Fix typo

* Add blank line between directives and actual code

* Re-format files

* Remove type in comment

---------

Co-authored-by: carlushuang <carlus.huang@amd.com>
Co-authored-by: rocking <ChunYu.Lai@amd.com>

[ROCm/composable_kernel commit: c156989298]

include/ck_tile/core/numeric/math.hpp

@@ -536,13 +536,20 @@ float log(float x) { return __logf(x); };
 CK_TILE_HOST
 float log(float x) { return std::logf(x); };
 
-CK_TILE_DEVICE uint32_t sad(uint32_t x, uint32_t y, uint32_t acc)
+CK_TILE_DEVICE uint16_t sad_u16(uint16_t x, uint16_t y, uint16_t acc)
 {
-    // TODO: this is hacky, we use u16
     return __builtin_amdgcn_sad_u16(x, y, acc);
 }
 
-CK_TILE_HOST uint32_t sad(uint32_t x, uint32_t y, uint32_t acc)
+CK_TILE_DEVICE uint32_t sad_u32(uint32_t x, uint32_t y, uint32_t acc)
+{
+    /// TODO: replace inline asm when intrinsic is available
+    uint32_t res;
+    asm volatile("v_sad_u32 %0, %1, %2, %3" : "=v"(res) : "v"(x), "v"(y), "v"(acc));
+    return res;
+}
+
+CK_TILE_HOST uint32_t sad_u32(uint32_t x, uint32_t y, uint32_t acc)
 {
     return (x > y ? (x - y) : (y - x)) + acc;
 }

include/ck_tile/core/tensor/tile_window.hpp

@@ -214,6 +214,12 @@ struct tile_window_with_static_distribution
 
     CK_TILE_DEVICE constexpr auto get_window_origin() const { return window_origin_; }
 
+    CK_TILE_DEVICE constexpr void
+    set_bottom_tensor_view_data_ptr(typename BottomTensorView::DataType* data)
+    {
+        bottom_tensor_view_.buf_.p_data_ = data;
+    }
+
     // move thread's window adaptor coordinate and bottom tensor coordinate
     // [p0, p1, ..., y0, y1, ...] ==> [x0, x1, ...] ==> [x0', x1', ...] ==> [offset]
     CK_TILE_DEVICE void move_window_adaptor_and_bottom_tensor_thread_coordinate(
@@ -843,6 +849,17 @@ struct tile_window_with_static_lengths
 
     CK_TILE_DEVICE constexpr auto get_window_origin() const { return window_origin_; }
 
+    CK_TILE_DEVICE void set_window_origin(const BottomTensorIndex& new_window_origin)
+    {
+        window_origin_ = new_window_origin;
+    }
+
+    CK_TILE_DEVICE constexpr void
+    set_bottom_tensor_view_data_ptr(typename BottomTensorView::DataType* data)
+    {
+        bottom_tensor_view_.buf_.p_data_ = data;
+    }
+
     // move window-origin
     CK_TILE_DEVICE void move(const BottomTensorIndex& step) { window_origin_ += step; }
 
@@ -871,6 +888,39 @@ make_tile_window(const TensorView_& tensor_view,
         tensor_view, window_lengths, origin};
 }
 
+// duplicate tile window and replace its origin
+template <typename TensorView, typename WindowLengths>
+CK_TILE_DEVICE constexpr auto
+make_tile_window(const tile_window_with_static_lengths<TensorView, WindowLengths>& tile_window,
+                 const multi_index<TensorView::get_num_of_dimension()>& origin)
+{
+    return tile_window_with_static_lengths<TensorView, WindowLengths>{
+        tile_window.get_bottom_tensor_view(), tile_window.get_window_lengths(), origin};
+}
+
+template <typename TensorView, typename WindowLengths, typename StaticTileDistribution>
+CK_TILE_DEVICE constexpr auto
+make_tile_window(const tile_window_with_static_lengths<TensorView, WindowLengths>& tile_window,
+                 const multi_index<TensorView::get_num_of_dimension()>& origin,
+                 const StaticTileDistribution& tile_distribution)
+{
+    return make_tile_window(tile_window.get_bottom_tensor_view(),
+                            tile_window.get_window_lengths(),
+                            origin,
+                            tile_distribution);
+}
+
+template <typename TensorView, typename WindowLengths, typename StaticTileDistribution>
+CK_TILE_DEVICE constexpr auto
+make_tile_window(const tile_window_with_static_lengths<TensorView, WindowLengths>& tile_window,
+                 const StaticTileDistribution& tile_distribution)
+{
+    return make_tile_window(tile_window.get_bottom_tensor_view(),
+                            tile_window.get_window_lengths(),
+                            tile_window.get_window_origin(),
+                            tile_distribution);
+}
+
 template <typename TensorView_, typename WindowLengths_>
 CK_TILE_DEVICE void move_tile_window(
     tile_window_with_static_lengths<TensorView_, WindowLengths_>& window,

include/ck_tile/core/utility/type_traits.hpp

@@ -22,6 +22,23 @@ using remove_cvref_t = remove_cv_t<std::remove_reference_t<T>>;
 template <typename T>
 using remove_pointer_t = typename std::remove_pointer<T>::type;
 
+template <typename From, typename To>
+struct copy_const
+{
+    static_assert(!std::is_const_v<From>);
+
+    using type = To;
+};
+
+template <typename From, typename To>
+struct copy_const<const From, To>
+{
+    using type = std::add_const_t<typename copy_const<From, To>::type>;
+};
+
+template <typename From, typename To>
+using copy_const_t = typename copy_const<From, To>::type;
+
 namespace detail {
 template <class Default, class AlwaysVoid, template <class...> class Op, class... Args>
 struct detector