numba-cuda 0.8.0__py3-none-any.whl → 0.9.0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (227) hide show
  1. _numba_cuda_redirector.py +17 -13
  2. numba_cuda/VERSION +1 -1
  3. numba_cuda/_version.py +4 -1
  4. numba_cuda/numba/cuda/__init__.py +6 -2
  5. numba_cuda/numba/cuda/api.py +129 -86
  6. numba_cuda/numba/cuda/api_util.py +3 -3
  7. numba_cuda/numba/cuda/args.py +12 -16
  8. numba_cuda/numba/cuda/cg.py +6 -6
  9. numba_cuda/numba/cuda/codegen.py +74 -43
  10. numba_cuda/numba/cuda/compiler.py +232 -113
  11. numba_cuda/numba/cuda/cpp_function_wrappers.cu +1 -2
  12. numba_cuda/numba/cuda/cuda_fp16.h +661 -661
  13. numba_cuda/numba/cuda/cuda_fp16.hpp +3 -3
  14. numba_cuda/numba/cuda/cuda_paths.py +291 -99
  15. numba_cuda/numba/cuda/cudadecl.py +125 -69
  16. numba_cuda/numba/cuda/cudadrv/__init__.py +3 -1
  17. numba_cuda/numba/cuda/cudadrv/devicearray.py +185 -135
  18. numba_cuda/numba/cuda/cudadrv/devices.py +16 -11
  19. numba_cuda/numba/cuda/cudadrv/driver.py +463 -297
  20. numba_cuda/numba/cuda/cudadrv/drvapi.py +241 -207
  21. numba_cuda/numba/cuda/cudadrv/dummyarray.py +66 -54
  22. numba_cuda/numba/cuda/cudadrv/enums.py +1 -1
  23. numba_cuda/numba/cuda/cudadrv/error.py +6 -2
  24. numba_cuda/numba/cuda/cudadrv/libs.py +67 -63
  25. numba_cuda/numba/cuda/cudadrv/linkable_code.py +16 -1
  26. numba_cuda/numba/cuda/cudadrv/mappings.py +16 -14
  27. numba_cuda/numba/cuda/cudadrv/nvrtc.py +138 -29
  28. numba_cuda/numba/cuda/cudadrv/nvvm.py +296 -161
  29. numba_cuda/numba/cuda/cudadrv/rtapi.py +1 -1
  30. numba_cuda/numba/cuda/cudadrv/runtime.py +20 -8
  31. numba_cuda/numba/cuda/cudaimpl.py +317 -233
  32. numba_cuda/numba/cuda/cudamath.py +1 -1
  33. numba_cuda/numba/cuda/debuginfo.py +8 -6
  34. numba_cuda/numba/cuda/decorators.py +75 -45
  35. numba_cuda/numba/cuda/descriptor.py +1 -1
  36. numba_cuda/numba/cuda/device_init.py +69 -18
  37. numba_cuda/numba/cuda/deviceufunc.py +143 -98
  38. numba_cuda/numba/cuda/dispatcher.py +300 -213
  39. numba_cuda/numba/cuda/errors.py +13 -10
  40. numba_cuda/numba/cuda/extending.py +1 -1
  41. numba_cuda/numba/cuda/initialize.py +5 -3
  42. numba_cuda/numba/cuda/intrinsic_wrapper.py +3 -3
  43. numba_cuda/numba/cuda/intrinsics.py +31 -27
  44. numba_cuda/numba/cuda/kernels/reduction.py +13 -13
  45. numba_cuda/numba/cuda/kernels/transpose.py +3 -6
  46. numba_cuda/numba/cuda/libdevice.py +317 -317
  47. numba_cuda/numba/cuda/libdeviceimpl.py +3 -2
  48. numba_cuda/numba/cuda/locks.py +16 -0
  49. numba_cuda/numba/cuda/mathimpl.py +62 -57
  50. numba_cuda/numba/cuda/models.py +1 -5
  51. numba_cuda/numba/cuda/nvvmutils.py +103 -88
  52. numba_cuda/numba/cuda/printimpl.py +9 -5
  53. numba_cuda/numba/cuda/random.py +46 -36
  54. numba_cuda/numba/cuda/reshape_funcs.cu +1 -1
  55. numba_cuda/numba/cuda/runtime/__init__.py +1 -1
  56. numba_cuda/numba/cuda/runtime/memsys.cu +1 -1
  57. numba_cuda/numba/cuda/runtime/memsys.cuh +1 -1
  58. numba_cuda/numba/cuda/runtime/nrt.cu +3 -3
  59. numba_cuda/numba/cuda/runtime/nrt.py +48 -43
  60. numba_cuda/numba/cuda/simulator/__init__.py +22 -12
  61. numba_cuda/numba/cuda/simulator/api.py +38 -22
  62. numba_cuda/numba/cuda/simulator/compiler.py +2 -2
  63. numba_cuda/numba/cuda/simulator/cudadrv/__init__.py +8 -2
  64. numba_cuda/numba/cuda/simulator/cudadrv/devicearray.py +63 -55
  65. numba_cuda/numba/cuda/simulator/cudadrv/devices.py +13 -11
  66. numba_cuda/numba/cuda/simulator/cudadrv/driver.py +5 -5
  67. numba_cuda/numba/cuda/simulator/cudadrv/drvapi.py +2 -2
  68. numba_cuda/numba/cuda/simulator/cudadrv/libs.py +1 -1
  69. numba_cuda/numba/cuda/simulator/cudadrv/nvvm.py +3 -3
  70. numba_cuda/numba/cuda/simulator/cudadrv/runtime.py +3 -3
  71. numba_cuda/numba/cuda/simulator/kernel.py +43 -34
  72. numba_cuda/numba/cuda/simulator/kernelapi.py +31 -26
  73. numba_cuda/numba/cuda/simulator/reduction.py +1 -0
  74. numba_cuda/numba/cuda/simulator/vector_types.py +13 -9
  75. numba_cuda/numba/cuda/simulator_init.py +2 -4
  76. numba_cuda/numba/cuda/stubs.py +139 -102
  77. numba_cuda/numba/cuda/target.py +64 -47
  78. numba_cuda/numba/cuda/testing.py +24 -19
  79. numba_cuda/numba/cuda/tests/__init__.py +14 -12
  80. numba_cuda/numba/cuda/tests/cudadrv/test_array_attr.py +16 -17
  81. numba_cuda/numba/cuda/tests/cudadrv/test_context_stack.py +7 -7
  82. numba_cuda/numba/cuda/tests/cudadrv/test_cuda_array_slicing.py +73 -54
  83. numba_cuda/numba/cuda/tests/cudadrv/test_cuda_auto_context.py +1 -1
  84. numba_cuda/numba/cuda/tests/cudadrv/test_cuda_devicerecord.py +48 -50
  85. numba_cuda/numba/cuda/tests/cudadrv/test_cuda_driver.py +47 -29
  86. numba_cuda/numba/cuda/tests/cudadrv/test_cuda_libraries.py +3 -3
  87. numba_cuda/numba/cuda/tests/cudadrv/test_cuda_memory.py +19 -19
  88. numba_cuda/numba/cuda/tests/cudadrv/test_cuda_ndarray.py +108 -103
  89. numba_cuda/numba/cuda/tests/cudadrv/test_deallocations.py +20 -11
  90. numba_cuda/numba/cuda/tests/cudadrv/test_detect.py +20 -17
  91. numba_cuda/numba/cuda/tests/cudadrv/test_emm_plugins.py +8 -6
  92. numba_cuda/numba/cuda/tests/cudadrv/test_events.py +1 -1
  93. numba_cuda/numba/cuda/tests/cudadrv/test_host_alloc.py +8 -7
  94. numba_cuda/numba/cuda/tests/cudadrv/test_init.py +13 -13
  95. numba_cuda/numba/cuda/tests/cudadrv/test_inline_ptx.py +12 -9
  96. numba_cuda/numba/cuda/tests/cudadrv/test_linker.py +36 -31
  97. numba_cuda/numba/cuda/tests/cudadrv/test_managed_alloc.py +8 -7
  98. numba_cuda/numba/cuda/tests/cudadrv/test_module_callbacks.py +294 -0
  99. numba_cuda/numba/cuda/tests/cudadrv/test_mvc.py +10 -7
  100. numba_cuda/numba/cuda/tests/cudadrv/test_nvjitlink.py +24 -15
  101. numba_cuda/numba/cuda/tests/cudadrv/test_nvvm_driver.py +43 -41
  102. numba_cuda/numba/cuda/tests/cudadrv/test_pinned.py +4 -5
  103. numba_cuda/numba/cuda/tests/cudadrv/test_profiler.py +2 -2
  104. numba_cuda/numba/cuda/tests/cudadrv/test_ptds.py +28 -17
  105. numba_cuda/numba/cuda/tests/cudadrv/test_reset_device.py +1 -2
  106. numba_cuda/numba/cuda/tests/cudadrv/test_runtime.py +22 -14
  107. numba_cuda/numba/cuda/tests/cudadrv/test_select_device.py +1 -1
  108. numba_cuda/numba/cuda/tests/cudadrv/test_streams.py +4 -3
  109. numba_cuda/numba/cuda/tests/cudapy/cache_usecases.py +10 -4
  110. numba_cuda/numba/cuda/tests/cudapy/cache_with_cpu_usecases.py +1 -0
  111. numba_cuda/numba/cuda/tests/cudapy/extensions_usecases.py +7 -6
  112. numba_cuda/numba/cuda/tests/cudapy/jitlink.ptx +0 -2
  113. numba_cuda/numba/cuda/tests/cudapy/recursion_usecases.py +1 -0
  114. numba_cuda/numba/cuda/tests/cudapy/test_alignment.py +6 -5
  115. numba_cuda/numba/cuda/tests/cudapy/test_array.py +52 -42
  116. numba_cuda/numba/cuda/tests/cudapy/test_array_args.py +5 -6
  117. numba_cuda/numba/cuda/tests/cudapy/test_array_methods.py +1 -1
  118. numba_cuda/numba/cuda/tests/cudapy/test_atomics.py +501 -304
  119. numba_cuda/numba/cuda/tests/cudapy/test_blackscholes.py +57 -21
  120. numba_cuda/numba/cuda/tests/cudapy/test_boolean.py +3 -3
  121. numba_cuda/numba/cuda/tests/cudapy/test_caching.py +50 -37
  122. numba_cuda/numba/cuda/tests/cudapy/test_casting.py +29 -24
  123. numba_cuda/numba/cuda/tests/cudapy/test_cffi.py +11 -6
  124. numba_cuda/numba/cuda/tests/cudapy/test_compiler.py +84 -50
  125. numba_cuda/numba/cuda/tests/cudapy/test_complex.py +144 -73
  126. numba_cuda/numba/cuda/tests/cudapy/test_complex_kernel.py +2 -2
  127. numba_cuda/numba/cuda/tests/cudapy/test_const_string.py +37 -27
  128. numba_cuda/numba/cuda/tests/cudapy/test_constmem.py +43 -45
  129. numba_cuda/numba/cuda/tests/cudapy/test_cooperative_groups.py +21 -14
  130. numba_cuda/numba/cuda/tests/cudapy/test_cuda_array_interface.py +60 -55
  131. numba_cuda/numba/cuda/tests/cudapy/test_cuda_jit_no_types.py +3 -2
  132. numba_cuda/numba/cuda/tests/cudapy/test_datetime.py +26 -22
  133. numba_cuda/numba/cuda/tests/cudapy/test_debug.py +29 -27
  134. numba_cuda/numba/cuda/tests/cudapy/test_debuginfo.py +31 -28
  135. numba_cuda/numba/cuda/tests/cudapy/test_device_func.py +52 -45
  136. numba_cuda/numba/cuda/tests/cudapy/test_dispatcher.py +55 -43
  137. numba_cuda/numba/cuda/tests/cudapy/test_enums.py +6 -7
  138. numba_cuda/numba/cuda/tests/cudapy/test_errors.py +30 -15
  139. numba_cuda/numba/cuda/tests/cudapy/test_exception.py +11 -12
  140. numba_cuda/numba/cuda/tests/cudapy/test_extending.py +19 -12
  141. numba_cuda/numba/cuda/tests/cudapy/test_fastmath.py +77 -66
  142. numba_cuda/numba/cuda/tests/cudapy/test_forall.py +5 -3
  143. numba_cuda/numba/cuda/tests/cudapy/test_freevar.py +5 -3
  144. numba_cuda/numba/cuda/tests/cudapy/test_frexp_ldexp.py +1 -1
  145. numba_cuda/numba/cuda/tests/cudapy/test_globals.py +3 -5
  146. numba_cuda/numba/cuda/tests/cudapy/test_gufunc.py +144 -126
  147. numba_cuda/numba/cuda/tests/cudapy/test_gufunc_scalar.py +23 -18
  148. numba_cuda/numba/cuda/tests/cudapy/test_gufunc_scheduling.py +16 -22
  149. numba_cuda/numba/cuda/tests/cudapy/test_idiv.py +1 -3
  150. numba_cuda/numba/cuda/tests/cudapy/test_inspect.py +29 -20
  151. numba_cuda/numba/cuda/tests/cudapy/test_intrinsics.py +147 -99
  152. numba_cuda/numba/cuda/tests/cudapy/test_ipc.py +50 -36
  153. numba_cuda/numba/cuda/tests/cudapy/test_iterators.py +1 -2
  154. numba_cuda/numba/cuda/tests/cudapy/test_lang.py +4 -4
  155. numba_cuda/numba/cuda/tests/cudapy/test_laplace.py +6 -6
  156. numba_cuda/numba/cuda/tests/cudapy/test_libdevice.py +24 -20
  157. numba_cuda/numba/cuda/tests/cudapy/test_lineinfo.py +36 -31
  158. numba_cuda/numba/cuda/tests/cudapy/test_localmem.py +13 -13
  159. numba_cuda/numba/cuda/tests/cudapy/test_mandel.py +13 -6
  160. numba_cuda/numba/cuda/tests/cudapy/test_math.py +83 -66
  161. numba_cuda/numba/cuda/tests/cudapy/test_matmul.py +1 -3
  162. numba_cuda/numba/cuda/tests/cudapy/test_minmax.py +19 -58
  163. numba_cuda/numba/cuda/tests/cudapy/test_montecarlo.py +4 -4
  164. numba_cuda/numba/cuda/tests/cudapy/test_multigpu.py +9 -7
  165. numba_cuda/numba/cuda/tests/cudapy/test_multiprocessing.py +9 -8
  166. numba_cuda/numba/cuda/tests/cudapy/test_multithreads.py +12 -10
  167. numba_cuda/numba/cuda/tests/cudapy/test_nondet.py +1 -1
  168. numba_cuda/numba/cuda/tests/cudapy/test_operator.py +180 -96
  169. numba_cuda/numba/cuda/tests/cudapy/test_optimization.py +5 -5
  170. numba_cuda/numba/cuda/tests/cudapy/test_overload.py +37 -18
  171. numba_cuda/numba/cuda/tests/cudapy/test_powi.py +7 -7
  172. numba_cuda/numba/cuda/tests/cudapy/test_print.py +9 -7
  173. numba_cuda/numba/cuda/tests/cudapy/test_py2_div_issue.py +1 -1
  174. numba_cuda/numba/cuda/tests/cudapy/test_random.py +15 -10
  175. numba_cuda/numba/cuda/tests/cudapy/test_record_dtype.py +88 -87
  176. numba_cuda/numba/cuda/tests/cudapy/test_recursion.py +12 -10
  177. numba_cuda/numba/cuda/tests/cudapy/test_reduction.py +26 -11
  178. numba_cuda/numba/cuda/tests/cudapy/test_retrieve_autoconverted_arrays.py +7 -10
  179. numba_cuda/numba/cuda/tests/cudapy/test_serialize.py +4 -6
  180. numba_cuda/numba/cuda/tests/cudapy/test_slicing.py +1 -1
  181. numba_cuda/numba/cuda/tests/cudapy/test_sm.py +10 -9
  182. numba_cuda/numba/cuda/tests/cudapy/test_sm_creation.py +62 -43
  183. numba_cuda/numba/cuda/tests/cudapy/test_stream_api.py +7 -3
  184. numba_cuda/numba/cuda/tests/cudapy/test_sync.py +7 -5
  185. numba_cuda/numba/cuda/tests/cudapy/test_transpose.py +18 -11
  186. numba_cuda/numba/cuda/tests/cudapy/test_ufuncs.py +111 -88
  187. numba_cuda/numba/cuda/tests/cudapy/test_userexc.py +2 -3
  188. numba_cuda/numba/cuda/tests/cudapy/test_vector_type.py +305 -130
  189. numba_cuda/numba/cuda/tests/cudapy/test_vectorize.py +33 -36
  190. numba_cuda/numba/cuda/tests/cudapy/test_vectorize_complex.py +5 -5
  191. numba_cuda/numba/cuda/tests/cudapy/test_vectorize_decor.py +16 -12
  192. numba_cuda/numba/cuda/tests/cudapy/test_vectorize_device.py +7 -7
  193. numba_cuda/numba/cuda/tests/cudapy/test_vectorize_scalar_arg.py +6 -7
  194. numba_cuda/numba/cuda/tests/cudapy/test_warning.py +31 -29
  195. numba_cuda/numba/cuda/tests/cudapy/test_warp_ops.py +31 -25
  196. numba_cuda/numba/cuda/tests/cudasim/test_cudasim_issues.py +19 -13
  197. numba_cuda/numba/cuda/tests/data/jitlink.cu +1 -1
  198. numba_cuda/numba/cuda/tests/data/jitlink.ptx +0 -2
  199. numba_cuda/numba/cuda/tests/doc_examples/test_cg.py +15 -8
  200. numba_cuda/numba/cuda/tests/doc_examples/test_cpu_gpu_compat.py +4 -7
  201. numba_cuda/numba/cuda/tests/doc_examples/test_ffi.py +14 -9
  202. numba_cuda/numba/cuda/tests/doc_examples/test_laplace.py +22 -18
  203. numba_cuda/numba/cuda/tests/doc_examples/test_matmul.py +7 -4
  204. numba_cuda/numba/cuda/tests/doc_examples/test_montecarlo.py +2 -0
  205. numba_cuda/numba/cuda/tests/doc_examples/test_random.py +8 -4
  206. numba_cuda/numba/cuda/tests/doc_examples/test_reduction.py +2 -1
  207. numba_cuda/numba/cuda/tests/doc_examples/test_sessionize.py +94 -19
  208. numba_cuda/numba/cuda/tests/doc_examples/test_vecadd.py +2 -2
  209. numba_cuda/numba/cuda/tests/nocuda/test_dummyarray.py +91 -62
  210. numba_cuda/numba/cuda/tests/nocuda/test_function_resolution.py +14 -5
  211. numba_cuda/numba/cuda/tests/nocuda/test_import.py +25 -25
  212. numba_cuda/numba/cuda/tests/nocuda/test_library_lookup.py +40 -40
  213. numba_cuda/numba/cuda/tests/nocuda/test_nvvm.py +12 -10
  214. numba_cuda/numba/cuda/tests/nrt/test_nrt.py +16 -20
  215. numba_cuda/numba/cuda/tests/nrt/test_nrt_refct.py +12 -10
  216. numba_cuda/numba/cuda/tests/test_binary_generation/generate_raw_ltoir.py +2 -2
  217. numba_cuda/numba/cuda/types.py +5 -2
  218. numba_cuda/numba/cuda/ufuncs.py +382 -362
  219. numba_cuda/numba/cuda/utils.py +2 -2
  220. numba_cuda/numba/cuda/vector_types.py +2 -2
  221. numba_cuda/numba/cuda/vectorizers.py +37 -32
  222. {numba_cuda-0.8.0.dist-info → numba_cuda-0.9.0.dist-info}/METADATA +1 -1
  223. numba_cuda-0.9.0.dist-info/RECORD +253 -0
  224. {numba_cuda-0.8.0.dist-info → numba_cuda-0.9.0.dist-info}/WHEEL +1 -1
  225. numba_cuda-0.8.0.dist-info/RECORD +0 -251
  226. {numba_cuda-0.8.0.dist-info → numba_cuda-0.9.0.dist-info}/licenses/LICENSE +0 -0
  227. {numba_cuda-0.8.0.dist-info → numba_cuda-0.9.0.dist-info}/top_level.txt +0 -0
@@ -6,5 +6,5 @@ from ctypes import c_int, POINTER
6
6
 
7
7
  API_PROTOTYPES = {
8
8
  # cudaError_t cudaRuntimeGetVersion ( int* runtimeVersion )
9
- 'cudaRuntimeGetVersion': (c_int, POINTER(c_int)),
9
+ "cudaRuntimeGetVersion": (c_int, POINTER(c_int)),
10
10
  }
@@ -21,6 +21,7 @@ class CudaRuntimeAPIError(CudaRuntimeError):
21
21
  """
22
22
  Raised when there is an error accessing a C API from the CUDA Runtime.
23
23
  """
24
+
24
25
  def __init__(self, code, msg):
25
26
  self.code = code
26
27
  self.msg = msg
@@ -44,11 +45,13 @@ class Runtime:
44
45
  _logger = make_logger()
45
46
 
46
47
  if config.DISABLE_CUDA:
47
- msg = ("CUDA is disabled due to setting NUMBA_DISABLE_CUDA=1 "
48
- "in the environment, or because CUDA is unsupported on "
49
- "32-bit systems.")
48
+ msg = (
49
+ "CUDA is disabled due to setting NUMBA_DISABLE_CUDA=1 "
50
+ "in the environment, or because CUDA is unsupported on "
51
+ "32-bit systems."
52
+ )
50
53
  raise CudaSupportError(msg)
51
- self.lib = open_cudalib('cudart')
54
+ self.lib = open_cudalib("cudart")
52
55
 
53
56
  self.is_initialized = True
54
57
 
@@ -76,9 +79,10 @@ class Runtime:
76
79
  def _wrap_api_call(self, fname, libfn):
77
80
  @functools.wraps(libfn)
78
81
  def safe_cuda_api_call(*args):
79
- _logger.debug('call runtime api: %s', libfn.__name__)
82
+ _logger.debug("call runtime api: %s", libfn.__name__)
80
83
  retcode = libfn(*args)
81
84
  self._check_error(fname, retcode)
85
+
82
86
  return safe_cuda_api_call
83
87
 
84
88
  def _check_error(self, fname, retcode):
@@ -125,11 +129,19 @@ class Runtime:
125
129
  def supported_versions(self):
126
130
  """A tuple of all supported CUDA toolkit versions. Versions are given in
127
131
  the form ``(major_version, minor_version)``."""
128
- if sys.platform not in ('linux', 'win32') or config.MACHINE_BITS != 64:
132
+ if sys.platform not in ("linux", "win32") or config.MACHINE_BITS != 64:
129
133
  # Only 64-bit Linux and Windows are supported
130
134
  return ()
131
- return ((11, 0), (11, 1), (11, 2), (11, 3), (11, 4), (11, 5), (11, 6),
132
- (11, 7))
135
+ return (
136
+ (11, 0),
137
+ (11, 1),
138
+ (11, 2),
139
+ (11, 3),
140
+ (11, 4),
141
+ (11, 5),
142
+ (11, 6),
143
+ (11, 7),
144
+ )
133
145
 
134
146
 
135
147
  runtime = Runtime()