FazBrowse GitHub Viewer | Trending |
URL:
| Home
Tools: [Download Repo ZIP]   [Original HTTPS Page]

Merge remote-tracking branch 'origin/master' into cpp · arrayfire/arrayfire@3815389 · GitHub

Repository navigation

Commit 3815389

Browse files
committed
Merge remote-tracking branch 'origin/master' into cpp
Conflicts: test/basic.cpp
2 parents 4c7d0dc + 9f8885c commit 3815389

7 files changed

Lines changed: 22 additions & 10 deletions

File tree

‎common/CLKernelToH.cmake‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -28,7 +28,7 @@ include(CMakeParseArguments)
2828
set(BIN2CPP_PROGRAM "bin2cpp")
2929

3030
function(CL_KERNEL_TO_H)
31-
cmake_parse_arguments(RTCS "" "VARNAME;EXTENSION;OUTPUT_DIR;TARGETS;NAMESPACE" "SOURCES" ${ARGN})
31+
cmake_parse_arguments(RTCS "" "VARNAME;EXTENSION;OUTPUT_DIR;TARGETS;NAMESPACE;EOF" "SOURCES" ${ARGN})
3232

3333
set(_output_files "")
3434
foreach(_input_file ${RTCS_SOURCES})
@@ -37,7 +37,7 @@ function(CL_KERNEL_TO_H)
3737
get_filename_component(var_name "${_input_file}" NAME)
3838
get_filename_component(_name_we "${_input_file}" NAME_WE)
3939

40-
set(_namespace, ${RTCS_NAMESPACE})
40+
set(_namespace "${RTCS_NAMESPACE}")
4141
string(REPLACE "." "_" var_name ${var_name})
4242

4343
set(_output_path "${CMAKE_CURRENT_BINARY_DIR}/${RTCS_OUTPUT_DIR}")
@@ -48,7 +48,7 @@ function(CL_KERNEL_TO_H)
4848
DEPENDS ${_input_file} ${BIN2CPP_PROGRAM}
4949
COMMAND ${CMAKE_COMMAND} -E make_directory "${_output_path}"
5050
COMMAND ${CMAKE_COMMAND} -E echo "\\#include \\<${_path}/${_name_we}.hpp\\>" >>"${_output_file}"
51-
COMMAND ${BIN2CPP_PROGRAM} --file ${_name} --namespace ${_namespace} --output ${_output_file} --name ${var_name}
51+
COMMAND ${BIN2CPP_PROGRAM} --file ${_name} --namespace ${_namespace} --output ${_output_file} --name ${var_name} --eof ${RTCS_EOF}
5252
WORKING_DIRECTORY "${_path}"
5353
COMMENT "Compiling ${_input_file} to C++ source"
5454
)

‎common/bin2cpp.cpp‎

Lines changed: 9 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -69,6 +69,7 @@ parse_options(const vector<string>& args) {
6969
options["--file"] = "";
7070
options["--output"] = "";
7171
options["--namespace"] = "";
72+
options["--eof"] = "";
7273

7374
//Parse Arguments
7475
string curr_opt;
@@ -112,6 +113,7 @@ parse_options(const vector<string>& args) {
112113

113114
int main(int argc, const char * const * const argv)
114115
{
116+
115117
vector<string> args(argv, argv+argc);
116118

117119
opt_t&& options = parse_options(args);
@@ -163,6 +165,13 @@ int main(int argc, const char * const * const argv)
163165
add_tabs(level);
164166
}
165167
}
168+
169+
if (options["--eof"].c_str()[0] == '1') {
170+
// Add end of file character
171+
cout << "0x0";
172+
char_cnt++;
173+
}
174+
166175
cout << "};\n";
167176
add_tabs(--level);
168177
cout << "static const size_t " << options["--name"] << "_len" << " = " << std::dec << char_cnt << ";\n";

‎src/backend/cuda/CMakeLists.txt‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -42,6 +42,7 @@ CL_KERNEL_TO_H(
4242
OUTPUT_DIR ${ptx_headers}
4343
TARGETS ptx_targets
4444
NAMESPACE "cuda"
45+
EOF "1"
4546
)
4647

4748
IF(${CMAKE_CXX_COMPILER_ID} STREQUAL "Clang" AND "${APPLE}")

‎src/backend/cuda/jit.cpp‎

Lines changed: 6 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -11,6 +11,7 @@
1111
#include <ptx_headers/numeric.hpp>
1212
#include <ptx_headers/trig.hpp>
1313
#include <ptx_headers/hyper.hpp>
14+
#include <ptx_headers/cast.hpp>
1415
#include <platform.hpp>
1516
#include <dispatch.hpp>
1617
#include <err_cuda.hpp>
@@ -244,12 +245,12 @@ static kc_entry_t compileKernel(const char *ker_name, string jit_ker)
244245
CU_CHECK(cuLinkAddData(linkState, CU_JIT_INPUT_PTX, (void*)ptx,
245246
ptx_size, ker_name, 0, NULL, NULL));
246247

247-
CU_CHECK(cuLinkAddData(linkState, CU_JIT_INPUT_PTX, (void*)trig_ptx,
248-
trig_ptx_len, "trig", 0, NULL, NULL));
249-
250248
CU_CHECK(cuLinkAddData(linkState, CU_JIT_INPUT_PTX, (void*)arith_ptx,
251249
arith_ptx_len, "arith", 0, NULL, NULL));
252250

251+
CU_CHECK(cuLinkAddData(linkState, CU_JIT_INPUT_PTX, (void*)cast_ptx,
252+
cast_ptx_len, "cast", 0, NULL, NULL));
253+
253254
CU_CHECK(cuLinkAddData(linkState, CU_JIT_INPUT_PTX, (void*)exp_ptx,
254255
exp_ptx_len, "exp", 0, NULL, NULL));
255256

@@ -262,6 +263,8 @@ static kc_entry_t compileKernel(const char *ker_name, string jit_ker)
262263
CU_CHECK(cuLinkAddData(linkState, CU_JIT_INPUT_PTX, (void*)numeric_ptx,
263264
numeric_ptx_len, "numeric", 0, NULL, NULL));
264265

266+
CU_CHECK(cuLinkAddData(linkState, CU_JIT_INPUT_PTX, (void*)trig_ptx,
267+
trig_ptx_len, "trig", 0, NULL, NULL));
265268

266269
void *cubin;
267270
size_t cubinSize;

‎src/backend/cuda/math.hpp‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -82,8 +82,8 @@ namespace cuda
8282
template<> __device__ char limit_min<char>() { return 0x80; }
8383
template<> __device__ float limit_max<float>() { return CUDART_INF_F; }
8484
template<> __device__ float limit_min<float>() { return -CUDART_INF_F; }
85-
template<> __device__ double limit_max<double>() { return CUDART_INF_F; }
86-
template<> __device__ double limit_min<double>() { return -CUDART_INF_F; }
85+
template<> __device__ double limit_max<double>() { return CUDART_INF; }
86+
template<> __device__ double limit_min<double>() { return -CUDART_INF; }
8787
#endif
8888

8989
#define upcast cuComplexFloatToDouble

‎src/backend/opencl/CMakeLists.txt‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -46,6 +46,7 @@ CL_KERNEL_TO_H(
4646
OUTPUT_DIR ${cl_kernel_headers}
4747
TARGETS cl_kernel_targets
4848
NAMESPACE "opencl"
49+
EOF "0"
4950
)
5051

5152
SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fPIC -pthread")

‎test/basic.cpp‎

Lines changed: 0 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -131,7 +131,6 @@ TEST(BasicTests, Additionf64f64)
131131
ASSERT_NEAR(0.0f, err, 1e-8);
132132
}
133133

134-
#ifndef AF_CUDA //FIXME: CUDA has problems with casting
135134
TEST(BasicTests, Additionf32f64)
136135
{
137136
static const int ndims = 2;
@@ -161,7 +160,6 @@ TEST(BasicTests, Additionf32f64)
161160
}
162161
ASSERT_NEAR(0.0f, err, 1e-8);
163162
}
164-
#endif
165163

166164
TEST(BasicArrayTests, constant10x10)
167165
{

0 commit comments

Comments
 (0)

Back | FazBrowse Home | New Git URL