def cl_reduce(function, output, input, shared, group_size, initial=0.0):
i = c_uint(0)
lid = clrt.get_local_id(0)
gid = clrt.get_group_id(0)
gsize = clrt.get_num_groups(0)
gs2 = group_size * 2
stride = gs2 * gsize
i = gid * gs2 + lid
shared[lid] = initial
while i < input.size:
shared[lid] = function(shared[lid], input[i])
shared[lid] = function(shared[lid], input[i + group_size])
i += stride
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
#The clyther compiler identifies this loop as a constant a
# unrolls this loop
for cgs in [512, 256, 128, 64, 32, 16, 8, 4, 2]:
#acts as a preprocessor define #if (group_size >= 512) etc.
if group_size >= cgs:
if lid < cgs / 2:
shared[lid] = function(shared[lid], shared[lid + cgs / 2])
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
if lid == 0:
output[gid] = shared[0]
def reduce_kernel(function, output, array, shared, group_size):
lid = clrt.get_local_id(0)
gid = clrt.get_group_id(0)
stride = group_size
i = c_uint(gid * group_size + lid)
igs = i + group_size
tmp = array[i]
if igs < array.size:
tmp = function(tmp, array[igs])
i += stride * 2
while i < array.size:
tmp = function(tmp, array[i])
i += stride
shared[lid] = tmp
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
#The clyther compiler identifies this loop as a constant a
# unrolls this loop
for cgs in [512, 256, 128, 64, 32, 16, 8, 4, 2]:
#acts as a preprocessor define #if (group_size >= 512) etc.
if group_size >= cgs:
if lid < cgs / 2:
shared[lid] = function(shared[lid], shared[lid + cgs / 2])
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
if lid == 0:
output[gid] = shared[0]
def cl_reduce(function, output, input, shared, group_size, initial=0.0):
i = c_uint(0)
lid = clrt.get_local_id(0)
gid = clrt.get_group_id(0)
gsize = clrt.get_num_groups(0)
gs2 = group_size * 2
stride = gs2 * gsize
i = gid * gs2 + lid
shared[lid] = initial
while i < input.size:
shared[lid] = function(shared[lid], input[i])
shared[lid] = function(shared[lid], input[i + group_size])
i += stride
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
#The clyther compiler identifies this loop as a constant a
# unrolls this loop
for cgs in [512 , 256, 128, 64, 32, 16, 8, 4, 2]:
#acts as a preprocessor define #if (group_size >= 512) etc.
if group_size >= cgs:
if lid < cgs / 2:
shared[lid] = function(shared[lid] , shared[lid + cgs / 2])
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
if lid == 0:
output[gid] = shared[0]
def reduce_kernel(function, output, array, shared, group_size):
lid = clrt.get_local_id(0)
gid = clrt.get_group_id(0)
stride = group_size
i = c_uint(gid * group_size + lid)
igs = i + group_size
tmp = array[i]
if igs < array.size:
tmp = function(tmp, array[igs])
i += stride*2
while i < array.size:
tmp = function(tmp, array[i])
i += stride
shared[lid] = tmp
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
#The clyther compiler identifies this loop as a constant a
# unrolls this loop
for cgs in [512 , 256, 128, 64, 32, 16, 8, 4, 2]:
#acts as a preprocessor define #if (group_size >= 512) etc.
if group_size >= cgs:
if lid < cgs / 2:
shared[lid] = function(shared[lid] , shared[lid + cgs / 2])
clrt.barrier(clrt.CLK_LOCAL_MEM_FENCE)
if lid == 0:
output[gid] = shared[0]
def setslice(a, value):
i = clrt.get_global_id(0)
a[i] = value
clrt.barrier(clrt.CLK_GLOBAL_MEM_FENCE)
def setslice(a, value):
i = clrt.get_global_id(0)
a[i] = value
clrt.barrier(clrt.CLK_GLOBAL_MEM_FENCE)
