Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
lbmpy
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Model registry
Operate
Environments
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Terms and privacy
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
pycodegen
lbmpy
Commits
fd16f1f4
Commit
fd16f1f4
authored
4 years ago
by
Michael Kuron
Browse files
Options
Downloads
Plain Diff
Merge branch 'Fix_LBMPeriodicityHandling' into 'master'
Fix lbm periodicity handling See merge request
pycodegen/lbmpy!55
parents
3e293a52
54241096
No related branches found
No related tags found
1 merge request
!55
Fix lbm periodicity handling
Pipeline
#29625
passed
4 years ago
Stage: test
Stage: deploy
Changes
2
Pipelines
1
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
lbmpy/advanced_streaming/communication.py
+14
-6
14 additions, 6 deletions
lbmpy/advanced_streaming/communication.py
lbmpy_tests/advanced_streaming/test_communication.py
+80
-10
80 additions, 10 deletions
lbmpy_tests/advanced_streaming/test_communication.py
with
94 additions
and
16 deletions
lbmpy/advanced_streaming/communication.py
+
14
−
6
View file @
fd16f1f4
import
itertools
from
pystencils
import
Field
,
Assignment
from
pystencils.slicing
import
shift_slice
,
get_slice_before_ghost_layer
,
normalize_slice
from
lbmpy.advanced_streaming.utility
import
is_inplace
,
get_accessor
,
numeric_index
,
\
...
...
@@ -59,14 +60,17 @@ def get_communication_slices(
Return the source and destination slices for periodicity handling or communication between blocks.
:param stencil: The stencil used by the LB method.
:param comm_stencil: The stencil defining the communication directions. If None, it will be set to stencil.
:param comm_stencil: The stencil defining the communication directions. If None, it will be set to the
full stencil (D2Q9 in 2D, D3Q27 in 3D, etc.).
:param streaming_pattern: The streaming pattern.
:param prev_timestep: Timestep after which communication is run.
:param ghost_layers: Number of ghost layers in each direction.
"""
dim
=
len
(
stencil
[
0
])
if
comm_stencil
is
None
:
comm_stencil
=
stencil
comm_stencil
=
itertools
.
product
(
*
([
-
1
,
0
,
1
]
for
_
in
range
(
dim
)))
pdfs
=
Field
.
create_generic
(
'
pdfs
'
,
spatial_dimensions
=
len
(
stencil
[
0
]),
index_shape
=
(
len
(
stencil
),))
write_accesses
=
get_accessor
(
streaming_pattern
,
prev_timestep
).
write
(
pdfs
,
stencil
)
...
...
@@ -168,8 +172,9 @@ class LBMPeriodicityHandling:
stencil
=
get_stencil
(
stencil
)
self
.
stencil
=
stencil
self
.
dim
=
len
(
stencil
[
0
])
self
.
dh
=
data_handling
target
=
data_handling
.
default_target
assert
target
in
[
'
cpu
'
,
'
gpu
'
,
'
opencl
'
]
...
...
@@ -184,13 +189,16 @@ class LBMPeriodicityHandling:
self
.
pycuda_direct_copy
=
target
==
'
gpu
'
and
pycuda_direct_copy
def
is_copy_direction
(
direction
):
s
=
0
for
d
,
p
in
zip
(
direction
,
periodicity
):
s
+=
abs
(
d
)
if
d
!=
0
and
not
p
:
return
False
return
True
return
s
!=
0
copy_directions
=
tuple
(
filter
(
is_copy_direction
,
stencil
[
1
:]))
full_stencil
=
itertools
.
product
(
*
([
-
1
,
0
,
1
]
for
_
in
range
(
self
.
dim
)))
copy_directions
=
tuple
(
filter
(
is_copy_direction
,
full_stencil
))
self
.
comm_slices
=
[]
timesteps
=
get_timesteps
(
streaming_pattern
)
for
timestep
in
timesteps
:
...
...
@@ -224,7 +232,7 @@ class LBMPeriodicityHandling:
for
src
,
dst
in
self
.
comm_slices
[
timestep
.
idx
]:
kernels
.
append
(
periodic_pdf_copy_kernel
(
pdf_field
,
src
,
dst
,
target
=
self
.
target
,
pdf_field
,
src
,
dst
,
target
=
self
.
target
,
opencl_queue
=
self
.
opencl_queue
,
opencl_ctx
=
self
.
opencl_ctx
))
return
kernels
...
...
This diff is collapsed.
Click to expand it.
lbmpy_tests/advanced_streaming/test_communication.py
+
80
−
10
View file @
fd16f1f4
import
pystencils
as
ps
import
numpy
as
np
from
lbmpy.stencils
import
get_stencil
from
pystencils.slicing
import
get_slice_before_ghost_layer
,
get_ghost_region_slice
from
lbmpy.advanced_streaming.communication
import
get_communication_slices
,
_fix_length_one_slices
from
lbmpy.creationfunctions
import
create_lb_update_rule
from
lbmpy.advanced_streaming.communication
import
get_communication_slices
,
_fix_length_one_slices
,
\
LBMPeriodicityHandling
from
lbmpy.advanced_streaming.utility
import
streaming_patterns
,
Timestep
import
pytest
@pytest.mark.parametrize
(
'
stencil
'
,
[
'
D2Q9
'
,
'
D3Q19
'
,
'
D3Q27
'
])
@pytest.mark.parametrize
(
'
stencil
'
,
[
'
D2Q9
'
,
'
D3Q15
'
,
'
D3Q19
'
,
'
D3Q27
'
])
@pytest.mark.parametrize
(
'
streaming_pattern
'
,
streaming_patterns
)
@pytest.mark.parametrize
(
'
timestep
'
,
[
Timestep
.
EVEN
,
Timestep
.
ODD
])
def
test_slices_not_empty
(
stencil
,
streaming_pattern
,
timestep
):
stencil
=
get_stencil
(
stencil
)
dim
=
len
(
stencil
[
0
])
q
=
len
(
stencil
)
arr
=
np
.
zeros
(
(
4
,)
*
dim
+
(
q
,)
)
slices
=
get_communication_slices
(
stencil
,
streaming_pattern
=
streaming_pattern
,
prev_timestep
=
timestep
,
ghost_layers
=
1
)
arr
=
np
.
zeros
((
4
,)
*
dim
+
(
q
,))
slices
=
get_communication_slices
(
stencil
,
streaming_pattern
=
streaming_pattern
,
prev_timestep
=
timestep
,
ghost_layers
=
1
)
for
_
,
slices_list
in
slices
.
items
():
for
src
,
dst
in
slices_list
:
assert
all
(
s
!=
0
for
s
in
arr
[
src
].
shape
)
assert
all
(
s
!=
0
for
s
in
arr
[
dst
].
shape
)
@pytest.mark.parametrize
(
'
stencil
'
,
[
'
D2Q9
'
,
'
D3Q19
'
,
'
D3Q27
'
])
@pytest.mark.parametrize
(
'
stencil
'
,
[
'
D2Q9
'
,
'
D3Q15
'
,
'
D3Q19
'
,
'
D3Q27
'
])
@pytest.mark.parametrize
(
'
streaming_pattern
'
,
streaming_patterns
)
@pytest.mark.parametrize
(
'
timestep
'
,
[
Timestep
.
EVEN
,
Timestep
.
ODD
])
def
test_src_dst_same_shape
(
stencil
,
streaming_pattern
,
timestep
):
stencil
=
get_stencil
(
stencil
)
dim
=
len
(
stencil
[
0
])
q
=
len
(
stencil
)
arr
=
np
.
zeros
(
(
4
,)
*
dim
+
(
q
,)
)
slices
=
get_communication_slices
(
stencil
,
streaming_pattern
=
streaming_pattern
,
prev_timestep
=
timestep
,
ghost_layers
=
1
)
arr
=
np
.
zeros
((
4
,)
*
dim
+
(
q
,))
slices
=
get_communication_slices
(
stencil
,
streaming_pattern
=
streaming_pattern
,
prev_timestep
=
timestep
,
ghost_layers
=
1
)
for
_
,
slices_list
in
slices
.
items
():
for
src
,
dst
in
slices_list
:
src_shape
=
arr
[
src
].
shape
...
...
@@ -37,7 +45,7 @@ def test_src_dst_same_shape(stencil, streaming_pattern, timestep):
assert
src_shape
==
dst_shape
@pytest.mark.parametrize
(
'
stencil
'
,
[
'
D2Q9
'
,
'
D3Q19
'
,
'
D3Q27
'
])
@pytest.mark.parametrize
(
'
stencil
'
,
[
'
D2Q9
'
,
'
D3Q15
'
,
'
D3Q19
'
,
'
D3Q27
'
])
def
test_pull_communication_slices
(
stencil
):
stencil
=
get_stencil
(
stencil
)
...
...
@@ -52,9 +60,71 @@ def test_pull_communication_slices(stencil):
src
=
s
[
0
][:
-
1
]
dst
=
s
[
1
][:
-
1
]
break
inner_slice
=
_fix_length_one_slices
(
get_slice_before_ghost_layer
(
d
,
ghost_layers
=
1
))
inv_dir
=
(
-
e
for
e
in
d
)
gl_slice
=
_fix_length_one_slices
(
get_ghost_region_slice
(
inv_dir
,
ghost_layers
=
1
))
assert
src
==
inner_slice
assert
dst
==
gl_slice
\ No newline at end of file
assert
dst
==
gl_slice
@pytest.mark.parametrize
(
'
stencil_name
'
,
[
'
D2Q9
'
,
'
D3Q15
'
,
'
D3Q19
'
,
'
D3Q27
'
])
def
test_optimised_and_full_communication_equivalence
(
stencil_name
):
target
=
'
cpu
'
stencil
=
get_stencil
(
stencil_name
)
dim
=
len
(
stencil
[
0
])
domain_size
=
(
4
,
)
*
dim
dh
=
ps
.
create_data_handling
(
domain_size
,
periodicity
=
(
True
,
)
*
dim
,
parallel
=
False
,
default_target
=
target
)
pdf
=
dh
.
add_array
(
"
pdf
"
,
values_per_cell
=
len
(
stencil
),
dtype
=
np
.
int64
)
dh
.
fill
(
"
pdf
"
,
0
,
ghost_layers
=
True
)
pdf_tmp
=
dh
.
add_array
(
"
pdf_tmp
"
,
values_per_cell
=
len
(
stencil
),
dtype
=
np
.
int64
)
dh
.
fill
(
"
pdf_tmp
"
,
0
,
ghost_layers
=
True
)
gl
=
dh
.
ghost_layers_of_field
(
"
pdf
"
)
num
=
0
for
idx
,
x
in
np
.
ndenumerate
(
dh
.
cpu_arrays
[
'
pdf
'
]):
dh
.
cpu_arrays
[
'
pdf
'
][
idx
]
=
num
dh
.
cpu_arrays
[
'
pdf_tmp
'
][
idx
]
=
num
num
+=
1
ac
=
create_lb_update_rule
(
stencil
=
stencil
,
optimization
=
{
"
symbolic_field
"
:
pdf
,
"
symbolic_temporary_field
"
:
pdf_tmp
},
kernel_type
=
'
stream_pull_only
'
)
ast
=
ps
.
create_kernel
(
ac
,
target
=
dh
.
default_target
,
cpu_openmp
=
True
)
stream
=
ast
.
compile
()
full_communication
=
dh
.
synchronization_function
(
pdf
.
name
,
target
=
dh
.
default_target
,
optimization
=
{
"
openmp
"
:
True
})
full_communication
()
dh
.
run_kernel
(
stream
)
dh
.
swap
(
"
pdf
"
,
"
pdf_tmp
"
)
pdf_full_communication
=
np
.
copy
(
dh
.
cpu_arrays
[
'
pdf
'
])
num
=
0
for
idx
,
x
in
np
.
ndenumerate
(
dh
.
cpu_arrays
[
'
pdf
'
]):
dh
.
cpu_arrays
[
'
pdf
'
][
idx
]
=
num
dh
.
cpu_arrays
[
'
pdf_tmp
'
][
idx
]
=
num
num
+=
1
optimised_communication
=
LBMPeriodicityHandling
(
stencil
=
stencil
,
data_handling
=
dh
,
pdf_field_name
=
pdf
.
name
,
streaming_pattern
=
'
pull
'
)
optimised_communication
()
dh
.
run_kernel
(
stream
)
dh
.
swap
(
"
pdf
"
,
"
pdf_tmp
"
)
if
dim
==
3
:
for
i
in
range
(
gl
,
domain_size
[
0
]):
for
j
in
range
(
gl
,
domain_size
[
1
]):
for
k
in
range
(
gl
,
domain_size
[
2
]):
for
f
in
range
(
len
(
stencil
)):
assert
dh
.
cpu_arrays
[
'
pdf
'
][
i
,
j
,
k
,
f
]
==
pdf_full_communication
[
i
,
j
,
k
,
f
],
print
(
f
)
else
:
for
i
in
range
(
gl
,
domain_size
[
0
]):
for
j
in
range
(
gl
,
domain_size
[
1
]):
for
f
in
range
(
len
(
stencil
)):
assert
dh
.
cpu_arrays
[
'
pdf
'
][
i
,
j
,
f
]
==
pdf_full_communication
[
i
,
j
,
f
]
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment