Replace RMM CUDA Python bindings with those provided by CUDA-Python (#…

…451)

As a follow up to rapidsai/rmm#930, fix RAFT to rely on CUDA Python directly rather than custom  CUDA bindings that RMM provided.

Authors:
  - Ashwin Srinath (https://github.com/shwina)

Approvers:
  - Corey J. Nolet (https://github.com/cjnolet)
  - Jordan Jacobelli (https://github.com/Ethyling)

URL: #451

conda/environments/raft_dev_cuda11.5.yml

@@ -6,6 +6,7 @@ channels:
 - conda-forge
 dependencies:
 - cudatoolkit=11.5
+- cuda-python >=11.5,<12.0
 - clang=11.1.0
 - clang-tools=11.1.0
 - rapids-build-env=22.02.*

python/raft/common/cuda.pxd

@@ -1,5 +1,5 @@
 #
-# Copyright (c) 2019, NVIDIA CORPORATION.
+# Copyright (c) 2019-2022, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -14,23 +14,9 @@
 # limitations under the License.
 #
 
-# cython: profile=False
-# distutils: language = c++
-# cython: embedsignature = True
-# cython: language_level = 3
+from cuda.ccudart cimport cudaStream_t
 
+cdef class Stream:
+    cdef cudaStream_t s
 
-# Populate this with more typedef's (eg: events) as and when needed
-cdef extern from * nogil:
-    ctypedef void* _Stream "cudaStream_t"
-    ctypedef int   _Error  "cudaError_t"
-
-
-# Populate this with more runtime api method declarations as and when needed
-cdef extern from "cuda_runtime_api.h" nogil:
-    _Error cudaStreamCreate(_Stream* s)
-    _Error cudaStreamDestroy(_Stream s)
-    _Error cudaStreamSynchronize(_Stream s)
-    _Error cudaGetLastError()
-    const char* cudaGetErrorString(_Error e)
-    const char* cudaGetErrorName(_Error e)
+    cdef cudaStream_t getStream(self)

python/raft/common/cuda.pyx

@@ -1,5 +1,5 @@
 #
-# Copyright (c) 2020-2021, NVIDIA CORPORATION.
+# Copyright (c) 2020-2022, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -19,10 +19,22 @@
 # cython: embedsignature = True
 # cython: language_level = 3
 
+from cuda.ccudart cimport(
+    cudaStream_t,
+    cudaError_t,
+    cudaSuccess,
+    cudaStreamCreate,
+    cudaStreamDestroy,
+    cudaStreamSynchronize,
+    cudaGetLastError,
+    cudaGetErrorString,
+    cudaGetErrorName
+)
+
 
 class CudaRuntimeError(RuntimeError):
     def __init__(self, extraMsg=None):
-        cdef _Error e = cudaGetLastError()
+        cdef cudaError_t e = cudaGetLastError()
         cdef bytes errMsg = cudaGetErrorString(e)
         cdef bytes errName = cudaGetErrorName(e)
         msg = "Error! %s reason='%s'" % (errName.decode(), errMsg.decode())
@@ -45,29 +57,17 @@ cdef class Stream:
         stream.sync()
         del stream  # optional!
     """
-
-    # NOTE:
-    # If we store _Stream directly, this always leads to the following error:
-    #   "Cannot convert Python object to '_Stream'"
-    # I was unable to find a good solution to this in reasonable time. Also,
-    # since cudaStream_t is a pointer anyways, storing it as an integer should
-    # be just fine (although, that certainly is ugly and hacky!).
-    cdef size_t s
-
     def __cinit__(self):
-        if self.s != 0:
-            return
-        cdef _Stream stream
-        cdef _Error e = cudaStreamCreate(&stream)
-        if e != 0:
+        cdef cudaStream_t stream
+        cdef cudaError_t e = cudaStreamCreate(&stream)
+        if e != cudaSuccess:
             raise CudaRuntimeError("Stream create")
-        self.s = <size_t>stream
+        self.s = stream
 
     def __dealloc__(self):
         self.sync()
-        cdef _Stream stream = <_Stream>self.s
-        cdef _Error e = cudaStreamDestroy(stream)
-        if e != 0:
+        cdef cudaError_t e = cudaStreamDestroy(self.s)
+        if e != cudaSuccess:
             raise CudaRuntimeError("Stream destroy")
 
     def sync(self):
@@ -76,10 +76,9 @@ cdef class Stream:
         could raise exception due to issues with previous asynchronous
         launches
         """
-        cdef _Stream stream = <_Stream>self.s
-        cdef _Error e = cudaStreamSynchronize(stream)
-        if e != 0:
+        cdef cudaError_t e = cudaStreamSynchronize(self.s)
+        if e != cudaSuccess:
             raise CudaRuntimeError("Stream sync")
 
-    def getStream(self):
+    cdef cudaStream_t getStream(self):
         return self.s

python/raft/common/handle.pxd

@@ -1,5 +1,5 @@
 #
-# Copyright (c) 2020, NVIDIA CORPORATION.
+# Copyright (c) 2020-2022, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -21,7 +21,6 @@
 
 
 from libcpp.memory cimport shared_ptr
-from .cuda cimport _Stream
 from rmm._lib.cuda_stream_view cimport cuda_stream_view
 from rmm._lib.cuda_stream_pool cimport cuda_stream_pool
 from libcpp.memory cimport shared_ptr

python/raft/common/handle.pyx

@@ -1,5 +1,5 @@
 #
-# Copyright (c) 2020, NVIDIA CORPORATION.
+# Copyright (c) 2020-2022, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -24,9 +24,10 @@ from libcpp.memory cimport shared_ptr
 from rmm._lib.cuda_stream_view cimport cuda_stream_per_thread
 from rmm._lib.cuda_stream_view cimport cuda_stream_view
 
-from .cuda cimport _Stream, _Error, cudaStreamSynchronize
+from .cuda cimport Stream
 from .cuda import CudaRuntimeError
 
+
 cdef class Handle:
     """
     Handle is a lightweight python wrapper around the corresponding C++ class
@@ -51,7 +52,7 @@ cdef class Handle:
         del handle  # optional!
     """
 
-    def __cinit__(self, stream=None, n_streams=0):
+    def __cinit__(self, stream: Stream = None, n_streams=0):
         self.n_streams = n_streams
         if n_streams > 0:
             self.stream_pool.reset(new cuda_stream_pool(n_streams))
@@ -64,7 +65,7 @@ cdef class Handle:
                                           self.stream_pool))
         else:
             # this constructor constructs a handle on user stream
-            c_stream = cuda_stream_view(<_Stream><size_t> stream.getStream())
+            c_stream = cuda_stream_view(stream.getStream())
             self.c_obj.reset(new handle_t(c_stream,
                                           self.stream_pool))