Skip to content
Snippets Groups Projects
Commit 1754ef27 authored by Martin Bauer's avatar Martin Bauer
Browse files

CUDA indexing: clip to maximum cuda block size

- previous method did not work with kernels generated for walberla where
  block size changes are made at runtime
- device query does not always work, since the compile system may have
  no GPU or not the same GPU
-> max block size is passed as parameter and only optionally determined
   by a device query
parent a278f336
No related branches found
No related tags found
No related merge requests found
...@@ -9,8 +9,6 @@ from functools import partial ...@@ -9,8 +9,6 @@ from functools import partial
from pystencils.sympyextensions import prod, is_integer_sequence from pystencils.sympyextensions import prod, is_integer_sequence
AUTO_BLOCK_SIZE_LIMITING = False
BLOCK_IDX = [TypedSymbol("blockIdx." + coord, create_type("int")) for coord in ('x', 'y', 'z')] BLOCK_IDX = [TypedSymbol("blockIdx." + coord, create_type("int")) for coord in ('x', 'y', 'z')]
THREAD_IDX = [TypedSymbol("threadIdx." + coord, create_type("int")) for coord in ('x', 'y', 'z')] THREAD_IDX = [TypedSymbol("threadIdx." + coord, create_type("int")) for coord in ('x', 'y', 'z')]
BLOCK_DIM = [TypedSymbol("blockDim." + coord, create_type("int")) for coord in ('x', 'y', 'z')] BLOCK_DIM = [TypedSymbol("blockDim." + coord, create_type("int")) for coord in ('x', 'y', 'z')]
...@@ -84,17 +82,26 @@ class BlockIndexing(AbstractIndexing): ...@@ -84,17 +82,26 @@ class BlockIndexing(AbstractIndexing):
compile_time_block_size: compile in concrete block size, otherwise the cuda variable 'blockDim' is used compile_time_block_size: compile in concrete block size, otherwise the cuda variable 'blockDim' is used
""" """
def __init__(self, field, iteration_slice, def __init__(self, field, iteration_slice,
block_size=(16, 16, 1), permute_block_size_dependent_on_layout=True, compile_time_block_size=False): block_size=(16, 16, 1), permute_block_size_dependent_on_layout=True, compile_time_block_size=False,
maximum_block_size=(1024, 1024, 64)):
if field.spatial_dimensions > 3: if field.spatial_dimensions > 3:
raise NotImplementedError("This indexing scheme supports at most 3 spatial dimensions") raise NotImplementedError("This indexing scheme supports at most 3 spatial dimensions")
if permute_block_size_dependent_on_layout: if permute_block_size_dependent_on_layout:
block_size = self.permute_block_size_according_to_layout(block_size, field.layout) block_size = self.permute_block_size_according_to_layout(block_size, field.layout)
if AUTO_BLOCK_SIZE_LIMITING:
block_size = self.limit_block_size_to_device_maximum(block_size)
self._block_size = block_size self._block_size = block_size
if maximum_block_size == 'auto':
# Get device limits
import pycuda.driver as cuda
# noinspection PyUnresolvedReferences
import pycuda.autoinit # NOQA
da = cuda.device_attribute
device = cuda.Context.get_device()
maximum_block_size = tuple(device.get_attribute(a)
for a in (da.MAX_BLOCK_DIM_X, da.MAX_BLOCK_DIM_Y, da.MAX_BLOCK_DIM_Z))
self._maximum_block_size = maximum_block_size
self._iterationSlice = normalize_slice(iteration_slice, field.spatial_shape) self._iterationSlice = normalize_slice(iteration_slice, field.spatial_shape)
self._dim = field.spatial_dimensions self._dim = field.spatial_dimensions
self._symbolic_shape = [e if isinstance(e, sp.Basic) else None for e in field.spatial_shape] self._symbolic_shape = [e if isinstance(e, sp.Basic) else None for e in field.spatial_shape]
...@@ -125,6 +132,7 @@ class BlockIndexing(AbstractIndexing): ...@@ -125,6 +132,7 @@ class BlockIndexing(AbstractIndexing):
adapted_block_size.append(sp.Min(block_size[i] * factor, widths[i])) adapted_block_size.append(sp.Min(block_size[i] * factor, widths[i]))
block_size = tuple(adapted_block_size) + extend_bs block_size = tuple(adapted_block_size) + extend_bs
block_size = tuple(sp.Min(bs, max_bs) for bs, max_bs in zip(block_size, self._maximum_block_size))
grid = tuple(div_ceil(length, block_size) for length, block_size in zip(widths, block_size)) grid = tuple(div_ceil(length, block_size) for length, block_size in zip(widths, block_size))
extend_gr = (1,) * (3 - len(grid)) extend_gr = (1,) * (3 - len(grid))
...@@ -140,66 +148,6 @@ class BlockIndexing(AbstractIndexing): ...@@ -140,66 +148,6 @@ class BlockIndexing(AbstractIndexing):
condition = sp.And(condition, c) condition = sp.And(condition, c)
return Block([Conditional(condition, kernel_content)]) return Block([Conditional(condition, kernel_content)])
@staticmethod
def limit_block_size_to_device_maximum(block_size):
"""Changes block size to match device limits.
* if the total amount of threads is too big for the current device, the biggest coordinate is divided by 2.
* next, if one component is still too big, the component which is too big is divided by 2 and the smallest
component is multiplied by 2, such that the total amount of threads stays the same
Returns:
the altered block_size
"""
# Get device limits
import pycuda.driver as cuda
# noinspection PyUnresolvedReferences
import pycuda.autoinit # NOQA
da = cuda.device_attribute
device = cuda.Context.get_device()
block_size = list(block_size)
max_threads = device.get_attribute(da.MAX_THREADS_PER_BLOCK)
max_block_size = [device.get_attribute(a)
for a in (da.MAX_BLOCK_DIM_X, da.MAX_BLOCK_DIM_Y, da.MAX_BLOCK_DIM_Z)]
def prod(seq):
result = 1
for e in seq:
result *= e
return result
def get_index_of_too_big_element():
for i, bs in enumerate(block_size):
if bs > max_block_size[i]:
return i
return None
def get_index_of_too_small_element():
for i, bs in enumerate(block_size):
if bs // 2 <= max_block_size[i]:
return i
return None
# Reduce the total number of threads if necessary
while prod(block_size) > max_threads:
item_to_reduce = block_size.index(max(block_size))
for j, block_size_entry in enumerate(block_size):
if block_size_entry > max_block_size[j]:
item_to_reduce = j
block_size[item_to_reduce] //= 2
# Cap individual elements
too_big_element_index = get_index_of_too_big_element()
while too_big_element_index is not None:
too_small_element_index = get_index_of_too_small_element()
block_size[too_small_element_index] *= 2
block_size[too_big_element_index] //= 2
too_big_element_index = get_index_of_too_big_element()
return tuple(block_size)
@staticmethod @staticmethod
def limit_block_size_by_register_restriction(block_size, required_registers_per_thread, device=None): def limit_block_size_by_register_restriction(block_size, required_registers_per_thread, device=None):
"""Shrinks the block_size if there are too many registers used per multiprocessor. """Shrinks the block_size if there are too many registers used per multiprocessor.
...@@ -307,6 +255,7 @@ class LineIndexing(AbstractIndexing): ...@@ -307,6 +255,7 @@ class LineIndexing(AbstractIndexing):
def symbolic_parameters(self): def symbolic_parameters(self):
return set() return set()
# -------------------------------------- Helper functions -------------------------------------------------------------- # -------------------------------------- Helper functions --------------------------------------------------------------
def _get_start_from_slice(iteration_slice): def _get_start_from_slice(iteration_slice):
......
...@@ -147,11 +147,6 @@ def test_periodicity(): ...@@ -147,11 +147,6 @@ def test_periodicity():
np.testing.assert_equal(cpu_result, gpu_result) np.testing.assert_equal(cpu_result, gpu_result)
def test_block_size_limiting():
res = BlockIndexing.limit_block_size_to_device_maximum((4096, 4096, 4096))
assert all(r < 4096 for r in res)
def test_block_indexing(): def test_block_indexing():
f = fields("f: [3D]") f = fields("f: [3D]")
bi = BlockIndexing(f, make_slice[:, :, :], block_size=(16, 8, 2), permute_block_size_dependent_on_layout=False) bi = BlockIndexing(f, make_slice[:, :, :], block_size=(16, 8, 2), permute_block_size_dependent_on_layout=False)
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment