Move USE_NEON declaration to cpu_check.h, include that header everywhere USE_NEON is used.
PiperOrigin-RevId: 254024542
This commit is contained in:
parent
87f2f24f95
commit
8c33948524
@ -426,6 +426,7 @@ cc_library(
|
||||
"//tensorflow/lite/c:c_api_internal",
|
||||
"//tensorflow/lite/kernels/internal:audio_utils",
|
||||
"//tensorflow/lite/kernels/internal:common",
|
||||
"//tensorflow/lite/kernels/internal:cpu_check",
|
||||
"//tensorflow/lite/kernels/internal:kernel_utils",
|
||||
"//tensorflow/lite/kernels/internal:optimized",
|
||||
"//tensorflow/lite/kernels/internal:optimized_base",
|
||||
|
||||
@ -16,6 +16,7 @@ limitations under the License.
|
||||
|
||||
#include "tensorflow/lite/c/builtin_op_data.h"
|
||||
#include "tensorflow/lite/c/c_api_internal.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/optimized_ops.h"
|
||||
#include "tensorflow/lite/kernels/internal/quantization_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/integer_ops/add.h"
|
||||
|
||||
@ -25,6 +25,7 @@ limitations under the License.
|
||||
#include "tensorflow/lite/c/builtin_op_data.h"
|
||||
#include "tensorflow/lite/c/c_api_internal.h"
|
||||
#include "tensorflow/lite/kernels/cpu_backend_support.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_multithread.h"
|
||||
#include "tensorflow/lite/kernels/internal/quantization_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/depthwiseconv_float.h"
|
||||
|
||||
@ -14,6 +14,7 @@ limitations under the License.
|
||||
==============================================================================*/
|
||||
#include "tensorflow/lite/c/builtin_op_data.h"
|
||||
#include "tensorflow/lite/c/c_api_internal.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/optimized_ops.h"
|
||||
#include "tensorflow/lite/kernels/internal/quantization_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/reference_ops.h"
|
||||
|
||||
@ -173,19 +173,10 @@ cc_library(
|
||||
hdrs = ["common.h"],
|
||||
copts = tflite_copts(),
|
||||
deps = [
|
||||
":cpu_check",
|
||||
":types",
|
||||
"@gemmlowp//:fixedpoint",
|
||||
] + select({
|
||||
":haswell": tflite_deps_intel,
|
||||
":ios_x86_64": tflite_deps_intel,
|
||||
":k8": tflite_deps_intel,
|
||||
":x86": tflite_deps_intel,
|
||||
":x86_64": tflite_deps_intel,
|
||||
":darwin": tflite_deps_intel,
|
||||
":darwin_x86_64": tflite_deps_intel,
|
||||
":freebsd": tflite_deps_intel,
|
||||
"//conditions:default": [],
|
||||
}),
|
||||
],
|
||||
)
|
||||
|
||||
cc_library(
|
||||
@ -669,8 +660,8 @@ cc_test(
|
||||
],
|
||||
shard_count = 2,
|
||||
deps = [
|
||||
":common",
|
||||
":compatibility",
|
||||
":cpu_check",
|
||||
":optimized_base",
|
||||
":reference_base",
|
||||
":test_util",
|
||||
@ -825,7 +816,17 @@ cc_library(
|
||||
],
|
||||
"//conditions:default": [],
|
||||
},
|
||||
),
|
||||
) + select({
|
||||
":haswell": tflite_deps_intel,
|
||||
":ios_x86_64": tflite_deps_intel,
|
||||
":k8": tflite_deps_intel,
|
||||
":x86": tflite_deps_intel,
|
||||
":x86_64": tflite_deps_intel,
|
||||
":darwin": tflite_deps_intel,
|
||||
":darwin_x86_64": tflite_deps_intel,
|
||||
":freebsd": tflite_deps_intel,
|
||||
"//conditions:default": [],
|
||||
}),
|
||||
)
|
||||
|
||||
cc_test(
|
||||
|
||||
@ -21,31 +21,8 @@ limitations under the License.
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#ifndef USE_NEON
|
||||
#if defined(__ARM_NEON__) || defined(__ARM_NEON)
|
||||
#define USE_NEON
|
||||
#include <arm_neon.h>
|
||||
#endif
|
||||
|
||||
#if defined __GNUC__ && defined __SSE4_1__ && !defined TF_LITE_DISABLE_X86_NEON
|
||||
#define USE_NEON
|
||||
|
||||
#define OPTIMIZED_OPS_H__IGNORE_DEPRECATED_DECLARATIONS
|
||||
#pragma GCC diagnostic push
|
||||
#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
|
||||
#pragma GCC diagnostic ignored "-Wattributes"
|
||||
|
||||
#pragma GCC diagnostic push
|
||||
#pragma GCC diagnostic ignored "-Wnarrowing"
|
||||
#pragma GCC diagnostic ignored "-Wsequence-point"
|
||||
|
||||
#include "NEON_2_SSE.h"
|
||||
|
||||
#pragma GCC diagnostic pop
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#include "fixedpoint/fixedpoint.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
namespace tflite {
|
||||
|
||||
@ -26,7 +26,7 @@ limitations under the License.
|
||||
#include <gtest/gtest.h>
|
||||
#include "tensorflow/lite/experimental/ruy/context.h"
|
||||
#include "tensorflow/lite/kernels/cpu_backend_context.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/test_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
|
||||
@ -17,6 +17,27 @@ limitations under the License.
|
||||
|
||||
#include "tensorflow/lite/kernels/cpu_backend_context.h"
|
||||
|
||||
#ifndef USE_NEON
|
||||
#if defined(__ARM_NEON__) || defined(__ARM_NEON)
|
||||
#define USE_NEON
|
||||
#include <arm_neon.h>
|
||||
#endif
|
||||
|
||||
#if defined __GNUC__ && defined __SSE4_1__ && !defined TF_LITE_DISABLE_X86_NEON
|
||||
#define USE_NEON
|
||||
|
||||
#pragma GCC diagnostic push
|
||||
#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
|
||||
#pragma GCC diagnostic ignored "-Wattributes"
|
||||
#pragma GCC diagnostic ignored "-Wnarrowing"
|
||||
#pragma GCC diagnostic ignored "-Wsequence-point"
|
||||
|
||||
#include "NEON_2_SSE.h"
|
||||
|
||||
#pragma GCC diagnostic pop
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#if !defined(NEON_OR_PORTABLE_USE_PORTABLE) && \
|
||||
!defined(NEON_OR_PORTABLE_USE_NEON)
|
||||
// If neither is defined, figure out if we can use NEON_OR_PORTABLE_USE_PORTABLE
|
||||
|
||||
@ -16,7 +16,7 @@ limitations under the License.
|
||||
#define TENSORFLOW_LITE_KERNELS_INTERNAL_OPTIMIZED_DEPTHWISECONV_3X3_FILTER_COMMON_H_
|
||||
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/depthwiseconv_uint8.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
|
||||
@ -16,7 +16,6 @@ limitations under the License.
|
||||
#define TENSORFLOW_LITE_KERNELS_INTERNAL_OPTIMIZED_DEPTHWISECONV_FLOAT_H_
|
||||
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
|
||||
@ -18,7 +18,6 @@ limitations under the License.
|
||||
#include <type_traits>
|
||||
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_uint8_3x3_filter.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/depthwiseconv_uint8.h"
|
||||
|
||||
@ -18,7 +18,7 @@ limitations under the License.
|
||||
#include <memory>
|
||||
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_3x3_filter_common.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/depthwiseconv_uint8.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
@ -24,8 +24,8 @@ limitations under the License.
|
||||
|
||||
#include <algorithm>
|
||||
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/compatibility.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_uint8.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_uint8_3x3_filter.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/depthwiseconv_uint8.h"
|
||||
|
||||
@ -17,6 +17,7 @@ limitations under the License.
|
||||
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
namespace tflite {
|
||||
|
||||
@ -18,7 +18,7 @@ limitations under the License.
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/cpu_backend_context.h"
|
||||
#include "tensorflow/lite/kernels/cpu_backend_threadpool.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_3x3_filter_common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/integer_ops/depthwise_conv_3x3_filter.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/depthwiseconv_uint8.h"
|
||||
|
||||
@ -18,7 +18,7 @@ limitations under the License.
|
||||
#include <memory>
|
||||
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_3x3_filter_common.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
|
||||
@ -17,6 +17,7 @@ limitations under the License.
|
||||
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/types.h"
|
||||
|
||||
namespace tflite {
|
||||
|
||||
@ -29,7 +29,7 @@ limitations under the License.
|
||||
|
||||
#include "fixedpoint/fixedpoint.h"
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/im2col_utils.h"
|
||||
#include "tensorflow/lite/kernels/internal/quantization_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/reference_ops.h"
|
||||
|
||||
@ -18,6 +18,7 @@ limitations under the License.
|
||||
#include "fixedpoint/fixedpoint.h"
|
||||
#include "profiling/instrumentation.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
|
||||
namespace tflite {
|
||||
namespace optimized_integer_ops {
|
||||
|
||||
@ -19,7 +19,6 @@ limitations under the License.
|
||||
#include <sys/types.h>
|
||||
|
||||
#include "public/gemmlowp.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/depthwiseconv_multithread.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/integer_ops/depthwise_conv.h"
|
||||
|
||||
@ -23,8 +23,8 @@ limitations under the License.
|
||||
|
||||
#include "tensorflow/lite/c/builtin_op_data.h"
|
||||
#include "tensorflow/lite/kernels/activation_functor.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/compatibility.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/neon_tensor_utils_impl.h"
|
||||
#include "tensorflow/lite/kernels/internal/round.h"
|
||||
|
||||
|
||||
@ -18,17 +18,12 @@ limitations under the License.
|
||||
// TODO(ghodrat): Remove this header file and the dependency to internal data
|
||||
// structure.
|
||||
#include "tensorflow/lite/c/builtin_op_data.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
|
||||
#if defined(_MSC_VER)
|
||||
#define __restrict__ __restrict
|
||||
#endif
|
||||
|
||||
#ifndef USE_NEON
|
||||
#if defined(__ARM_NEON__) || defined(__ARM_NEON)
|
||||
#define USE_NEON
|
||||
#endif // defined(__ARM_NEON__) || defined(__ARM_NEON)
|
||||
#endif // USE_NEON
|
||||
|
||||
namespace tflite {
|
||||
namespace tensor_utils {
|
||||
|
||||
|
||||
@ -42,7 +42,7 @@ limitations under the License.
|
||||
#include "tensorflow/lite/kernels/cpu_backend_gemm.h"
|
||||
#include "tensorflow/lite/kernels/cpu_backend_gemm_params.h"
|
||||
#include "tensorflow/lite/kernels/cpu_backend_threadpool.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/im2col_utils.h"
|
||||
#include "tensorflow/lite/kernels/internal/quantization_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/reference_ops.h"
|
||||
|
||||
@ -13,13 +13,8 @@ See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
==============================================================================*/
|
||||
#include "tensorflow/lite/kernels/internal/tensor_utils.h"
|
||||
#include "tensorflow/lite/kernels/internal/common.h"
|
||||
|
||||
#ifndef USE_NEON
|
||||
#if defined(__ARM_NEON__) || defined(__ARM_NEON)
|
||||
#define USE_NEON
|
||||
#endif // defined(__ARM_NEON__) || defined(__ARM_NEON)
|
||||
#endif // USE_NEON
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
|
||||
#ifdef USE_NEON
|
||||
#include "tensorflow/lite/kernels/internal/optimized/neon_tensor_utils.h"
|
||||
|
||||
@ -16,6 +16,7 @@ limitations under the License.
|
||||
|
||||
#include "tensorflow/lite/c/builtin_op_data.h"
|
||||
#include "tensorflow/lite/c/c_api_internal.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/optimized_ops.h"
|
||||
#include "tensorflow/lite/kernels/internal/quantization_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/integer_ops/mul.h"
|
||||
|
||||
@ -13,8 +13,10 @@ See the License for the specific language governing permissions and
|
||||
limitations under the License.
|
||||
==============================================================================*/
|
||||
#include <limits>
|
||||
|
||||
#include "tensorflow/lite/c/builtin_op_data.h"
|
||||
#include "tensorflow/lite/c/c_api_internal.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/cpu_check.h"
|
||||
#include "tensorflow/lite/kernels/internal/optimized/optimized_ops.h"
|
||||
#include "tensorflow/lite/kernels/internal/quantization_util.h"
|
||||
#include "tensorflow/lite/kernels/internal/reference/integer_ops/add.h"
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user