diff --git a/CMakeLists.txt b/CMakeLists.txt
index 273806eb0678b2c88d2268d321068e29433ae0e3..095c983cc1b98f676a263d91d501464cefdafaff 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -142,7 +142,7 @@ find_python_module(gmpy2 REQUIRED)
find_python_module(subprocess32 REQUIRED)
find_python_module(editdistance REQUIRED)
find_python_module(portalocker REQUIRED)
-find_python_module(colors.py REQUIRED)
+find_python_module(colors REQUIRED) # color.py package
find_python_module(tee REQUIRED)
find_python_module(primefac REQUIRED)
find_python_module(graph_tool REQUIRED)
diff --git a/ci/docker_images/ubuntu/bionic/Dockerfile b/ci/docker_images/ubuntu/bionic/Dockerfile
index 6284b2a57bf41209663b8496e350e02c203d4f16..575d705383d3c09d5fa9cc9a74f9b9e07b1aa6b6 100644
--- a/ci/docker_images/ubuntu/bionic/Dockerfile
+++ b/ci/docker_images/ubuntu/bionic/Dockerfile
@@ -146,8 +146,21 @@ RUN cd /tmp \
&& cd - \
&& rm -Rf /tmp/Oclgrind
+# clpeak
+RUN cd /tmp \
+ && git clone https://github.com/krrishnarraj/clpeak \
+ && cd clpeak/ \
+ && mkdir build \
+ && cd build/ \
+ && cmake .. \
+ && make \
+ && mv clpeak /usr/local/bin/ \
+ && cd - \
+ && rm -Rf /tmp/clpeak
+
# clFFT
RUN cd /tmp \
+ && ln -s /usr/local/lib /usr/local/lib64 \
&& git clone https://github.com/clMathLibraries/clFFT \
&& cd clFFT \
&& cd src \
@@ -180,7 +193,9 @@ RUN cd /tmp \
&& make install \
&& cd - \
&& rm -Rf /tmp/graph-tool-2.26
-
+
+# ensure all libraries are known by the runtime linker
+RUN ldconfig
# clean cached packages
RUN rm -rf /var/lib/apt/lists/*
diff --git a/examples/example_utils.py b/examples/example_utils.py
index b1b143088d62b3d2bf5b2815141b8bb834bd7f75..230667fff4deb041aabdb720ffe684a7acaf5d44 100644
--- a/examples/example_utils.py
+++ b/examples/example_utils.py
@@ -174,9 +174,9 @@ class HysopArgParser(argparse.ArgumentParser):
with self.redirect_stdout(rank, args):#, self.redirect_stderr(rank, args):
from hysop.tools.contexts import printoptions
- with printoptions(threshold=10000, linewidth=240,
+ with printoptions(threshold=10000, linewidth=2000,
nanstr='nan', infstr='inf',
- formatter={'float': lambda x: '{:>6.2f}'.format(x)}):
+ formatter={'float': lambda x: '{:>8.4f}'.format(x)}):
program(args, **kwds)
@staticmethod
diff --git a/examples/shear_layer/shear_layer.py b/examples/shear_layer/shear_layer.py
index 8693accf62e1b8ebc4f22d08469ede4d22069367..b463e5e8ab82885989b2a599419f1f1105df399c 100644
--- a/examples/shear_layer/shear_layer.py
+++ b/examples/shear_layer/shear_layer.py
@@ -161,15 +161,15 @@ def compute(args):
dfields = problem.input_discrete_fields
dfields[vorti].initialize(formula=init_vorticity)
- from hysop.tools.debug_utils import ImshowDebugger
- dbg = ImshowDebugger(data={'Wz':(dfields[vorti],0)},
- ntimes=1,
- enable_on_op_apply=False)
- dbg.synchronize_queue(cl_env.default_queue)
- dbg('initialization')
+ #from hysop.tools.debug_utils import ImshowDebugger
+ #dbg = ImshowDebugger(data={'Wz':(dfields[vorti],0)},
+ #ntimes=1,
+ #enable_on_op_apply=False)
+ #dbg.synchronize_queue(cl_env.default_queue)
+ #dbg('initialization')
# Finally solve the problem
- problem.solve(simu, dry_run=args.dry_run, dbg=dbg)
+ problem.solve(simu, dry_run=args.dry_run)
# Finalize
problem.finalize()
diff --git a/hysop/backend/device/codegen/functions/directional_remesh.py b/hysop/backend/device/codegen/functions/directional_remesh.py
index 5c4d042a485d70b2ad149065f8951f2307d1a1aa..87f00dbe22f09bbc9ce00dc05b2b4b080113f8cf 100644
--- a/hysop/backend/device/codegen/functions/directional_remesh.py
+++ b/hysop/backend/device/codegen/functions/directional_remesh.py
@@ -26,6 +26,9 @@ class DirectionalRemeshFunction(OpenClFunctionCodeGenerator):
use_short_circuit=None,
known_args=None,
debug_mode=False):
+
+
+ debug_mode = True
check_instance(sboundary,tuple,values=BoundaryCondition)
check_instance(remesh_kernel, RemeshKernel)
@@ -216,11 +219,11 @@ class DirectionalRemeshFunction(OpenClFunctionCodeGenerator):
printf_no_vector=True
if printf_no_vector:
- vnf = '[{}]'.format(', '.join('%2.2f' for _ in xrange(nparticles)))
+ vnf = '[{}]'.format(', '.join('%2.4f' for _ in xrange(nparticles)))
vni = '[{}]'.format(', '.join('%i' for _ in xrange(nparticles)))
expand_printf_vector = lambda x: ','.join(x[i] for i in xrange(nparticles))
else:
- vnf = '[%2.2{}f]'.format(self._printv(nparticles))
+ vnf = '[%2.4{}f]'.format(self._printv(nparticles))
vni = '[%{}i]'.format(self._printv(nparticles))
expand_printf_vector = lambda x: str(x)
epv = expand_printf_vector
@@ -264,7 +267,8 @@ class DirectionalRemeshFunction(OpenClFunctionCodeGenerator):
eps.declare(s)
s.jumpline()
- #s.decl_aligned_vars(tst0, tst1)
+ if debug_mode:
+ s.decl_aligned_vars(tst0, tst1)
s.decl_aligned_vars(rp, find, ind, y)
vone.declare(s, const=True)
s.decl_vars(*weights)
@@ -274,13 +278,14 @@ class DirectionalRemeshFunction(OpenClFunctionCodeGenerator):
s.append('{} = {}*{};'.format(rp, pos, inv_dx))
s.append('{} = fract({}, &{});'.format(y, rp, find))
s.append('{} = convert_{}_rtn({});'.format(ind, ivtype, find))
- #s.append('{} = floor({rp});'.format(tst0, rp=rp))
- #s.append('{} = {rp} - floor({rp});'.format(tst1, rp=rp))
if debug_mode:
- with self._ordered_wi_execution_(barrier=False):
- code = 'printf("%lu p={vnf}, rp={vnf}, floor(rp)={vnf}, rp-floor(rp)={vnf}, ind={vni}, y={vnf}, s0={vnf}, a=%f, b=%f, c=%f.\\n", {}, {}, {}, {}, {}, {}, {}, {}, {}, {}, {});'.format(
+ s.append('{} = floor({rp});'.format(tst0, rp=rp))
+ s.append('{} = {rp} - floor({rp});'.format(tst1, rp=rp))
+ if debug_mode:
+ #with self._ordered_wi_execution_(barrier=False):
+ with self._if_('(get_group_id(1)==0) && (get_group_id(1)==0) && get_local_id(0)==35'):
+ code = 'printf("%lu p={vnf}, p.invdx={vnf}, floor(p/dx)={vnf}, p/dx-floor(p/dx)={vnf}, ind={vni}, y={vnf}, s0={vnf}.\\n", {}, {}, {}, {}, {}, {}, {}, {});'.format(
'get_local_id(0)', epv(pos), epv(rp), epv(tst0), epv(tst1), epv(ind), epv(y), epv(scalars[0]),
- 'floor(4.5f)', 'floor((float2)(1.5f,2.54f)).x', 'floor((float2)(1.5f,2.54f)).y',
vnf=vnf, vni=vni)
s.append(code)
y.affect(s, init='{}-{}'.format(vone,y))
@@ -332,13 +337,14 @@ class DirectionalRemeshFunction(OpenClFunctionCodeGenerator):
val='{}*{}'.format(w,scalars[0])
cache_idx='{}+{}'.format(cache_ghosts, ind[0])
- if poly_splitted:
- printf = 'printf("BATCH {}: %lu wrote {vnf} at idx={vni} with Wl={vnf}, Wr={vnf}, W={vnf}, cond={vni}.\\n",{},{},{},{},{},{},{});'.format(
- i, 'get_local_id(0)', val, ind, wl, wr, w, 'y<0.5f', vnf=vnf, vni=vni)
- else:
- printf = 'printf("BATCH {}: %lu wrote {vnf} at idx {vni} with W={vnf}.\\n",{},{},{},{});'.format(
- i, 'get_local_id(0)', val, ind, w, vni=vni, vnf=vnf)
- s.append(printf)
+ with self._if_('(get_group_id(1)==0) && (get_group_id(1)==0) && get_local_id(0)==35'):
+ if poly_splitted:
+ printf = 'printf("BATCH {}: %lu wrote {vnf} at idx={vni} with Wl={vnf}, Wr={vnf}, W={vnf}, cond={vni}.\\n",{},{},{},{},{},{},{});'.format(
+ i, 'get_local_id(0)', val, ind, wl, wr, w, 'y<0.5f', vnf=vnf, vni=vni)
+ else:
+ printf = 'printf("BATCH {}: %lu wrote {vnf} at idx {vni} with W={vnf}.\\n",{},{},{},{});'.format(
+ i, 'get_local_id(0)', val, ind, w, vni=vni, vnf=vnf)
+ s.append(printf)
if not use_atomics:
s.barrier(_local=True)
if use_atomics:
diff --git a/hysop/backend/device/codegen/kernels/directional_remesh.py b/hysop/backend/device/codegen/kernels/directional_remesh.py
index 1aa60d7498c0410ba3c59f76e63c8db34843e7a4..63a4d1f6bd565b6098168131e21a149b5f4672aa 100644
--- a/hysop/backend/device/codegen/kernels/directional_remesh.py
+++ b/hysop/backend/device/codegen/kernels/directional_remesh.py
@@ -717,7 +717,7 @@ class DirectionalRemeshKernelGenerator(KernelCodeGenerator):
init=tg.dump(+0.0))
with s._else_():
end_id='{}+{}'.format(local_work, 'l')
- if debug_mode:
+ if debug_mode or True:
s.append('printf("%i loaded back %2.2f from position %i.\\n", {}, {}, {});'.format(local_id[0], cached_scalars[0][0][end_id], end_id))
with s._align_() as al:
for csi in cached_scalars:
@@ -725,6 +725,7 @@ class DirectionalRemeshKernelGenerator(KernelCodeGenerator):
csij.affect(al, align=True,
i='l',
init=csij[end_id])
+ s.barrier(_local=True)
s.jumpline()
diff --git a/hysop/backend/device/opencl/autotunable_kernels/remesh_dir.py b/hysop/backend/device/opencl/autotunable_kernels/remesh_dir.py
index 52bc6db03550148556395f674b216dbbdb9d4675..8c5da4faf68293ebb7f0524862b39166de715784 100644
--- a/hysop/backend/device/opencl/autotunable_kernels/remesh_dir.py
+++ b/hysop/backend/device/opencl/autotunable_kernels/remesh_dir.py
@@ -269,8 +269,8 @@ class OpenClAutotunableDirectionalRemeshKernel(OpenClAutotunableKernel):
precision = extra_kwds['precision']
force_atomics = extra_kwds['force_atomics']
- nparticles_options = [1,2,4,8,16]
- use_atomics_options = [True] if force_atomics else [True, False]
+ nparticles_options = [1,]#2,4,8,16]
+ use_atomics_options = [True,]# if force_atomics else [True, False]
if True in use_atomics_options:
cl_env = self.cl_env
msg=None
@@ -326,9 +326,9 @@ class OpenClAutotunableDirectionalRemeshKernel(OpenClAutotunableKernel):
cache_ghosts = extra_kwds['min_s_ghosts']
min_wg_size = npw.ones(shape=work_bounds.work_dim, dtype=npw.int32)
- min_wg_size[0] = max(min_wg_size[0], 2*cache_ghosts)
+ min_wg_size[0] = 36
max_wg_size = npw.ones(shape=work_bounds.work_dim, dtype=npw.int32)
- max_wg_size[0] = max(global_work_size[0], min_wg_size[0])
+ max_wg_size[0] = 36
return (min_wg_size, max_wg_size)
def compute_global_work_size(self, local_work_size, work, extra_parameters, extra_kwds):
diff --git a/hysop/backend/device/opencl/operator/directional/advection_dir.py b/hysop/backend/device/opencl/operator/directional/advection_dir.py
index f93666b08a51588279cb5af887aacae111a92a2c..c9352d1afc697fc24b6674dff838adde7c6650a5 100644
--- a/hysop/backend/device/opencl/operator/directional/advection_dir.py
+++ b/hysop/backend/device/opencl/operator/directional/advection_dir.py
@@ -13,7 +13,7 @@ from hysop.backend.device.opencl.opencl_kernel_launcher import OpenClKernelListL
class OpenClDirectionalAdvection(DirectionalAdvectionBase, OpenClDirectionalOperator):
DEBUG=False
- GRAPHICAL_DEBUG=False
+ GRAPHICAL_DEBUG=True
@debug
def __init__(self, force_atomics=False, relax_min_particles=False, remesh_criteria_eps=None,
@@ -190,8 +190,11 @@ class OpenClDirectionalAdvection(DirectionalAdvectionBase, OpenClDirectionalOper
elif self.GRAPHICAL_DEBUG:
import numpy as np
p = self.dposition
+ v = self.dvelocity
sin = self.dadvected_fields_in.values()[0]
sout = self.dadvected_fields_out.values()[0]
+
+ V = v[0].get()[v.compute_slices]
self.advection_kernel_launcher(queue=queue, dt=dt).wait()
@@ -206,21 +209,27 @@ class OpenClDirectionalAdvection(DirectionalAdvectionBase, OpenClDirectionalOper
S0 = np.sum(Sin, axis=1)
S1 = np.sum(Sout, axis=1)
- failed = np.where((S1-S0)>1e-1)[0]
+ failed = np.where((S1-S0)>1)[0]
if failed.size:
- for i in failed:
+ for i in failed[:1]:
print 'AT I={}'.format(i)
- slc = (i,slice(250,260))
+ slc = (i,slice(None))
+ print 'Vx'
+ print V[slc]
print 'P'
- print P[slc]*512
+ print P[slc]*P.shape[1]
print 'SIN'
print Sin[slc]
print 'SOUT'
print Sout[slc]
+ print 'INDEXES'
+ print np.arange(Sout[slc].size).astype(np.float32)
print 'SCALAR WAS NOT CONSERVED'
print ' Sin: {}'.format(S0[i]/S0.size)
print ' Sout: {}'.format(S1[i]/S1.size)
print
+ print 'FAILED INDICES'
+ print failed
import sys
sys.exit(1)
else: