feat(megengine): support tx2

GitOrigin-RevId: d1175a1f4a

dnn/test/cuda/accuracy_shake.cpp

@@ -96,8 +96,8 @@ TEST_F(CUDA, SHAKE_CONV_BIAS_FORWARD_QS8_NHWC) {
 }
 
 TEST_F(CUDA, SHAKE_CONV_BIAS_FORWARD_QS8_NCHWX) {
-    using Format = ConvBias::Param::Format;
     require_compute_capability(7, 5);
+    using Format = ConvBias::Param::Format;
     AccuracyShakeChecker<ConvBiasForward> checker(handle_cuda());
     UniformIntRNG int_rng{-5, 5};
     UniformFloatRNG float_rng{-50, 50};
@@ -135,6 +135,7 @@ TEST_F(CUDA, SHAKE_CONV_BIAS_FORWARD_QS8_NCHWX) {
 }
 
 TEST_F(CUDA, SHAKE_MATRIX_MUL_FORWARD) {
+    require_compute_capability(6, 1);
     AccuracyShakeChecker<MatrixMul> checker(handle_cuda());
 
     checker.set_dtype(0, dtype::Float32())
@@ -167,6 +168,7 @@ TEST_F(CUDA, SHAKE_BATCH_CONV_BIAS_QS8) {
 }
 
 TEST_F(CUDA, SHAKE_BATCHED_MATRIX_MUL) {
+    require_compute_capability(6, 1);
     AccuracyShakeChecker<BatchedMatrixMul> checker(handle_cuda());
 
     UniformIntRNG int_rng{-127, 127};
@@ -189,6 +191,7 @@ TEST_F(CUDA, SHAKE_BATCHED_MATRIX_MUL) {
 }
 
 TEST_F(CUDA, SHAKE_CONVOLUTION3D_FORWARD) {
+    require_compute_capability(6, 1);
     AccuracyShakeChecker<Convolution3DForward> checker(handle_cuda());
     NormalRNG default_rng;
     float scale = 1.0f / sqrt(5);
@@ -207,6 +210,7 @@ TEST_F(CUDA, SHAKE_CONVOLUTION3D_FORWARD) {
 }
 
 TEST_F(CUDA, SHAKE_LOCAL_SHARE) {
+    require_compute_capability(6, 1);
     AccuracyShakeChecker<LocalShare> checker(handle_cuda());
     using Param = LocalShare::Param;
     Param param;
@@ -216,6 +220,7 @@ TEST_F(CUDA, SHAKE_LOCAL_SHARE) {
 }
 
 TEST_F(CUDA, SHAKE_CONVOLUTION_BACKWARD_DATA) {
+    require_compute_capability(6, 1);
     AccuracyShakeChecker<ConvolutionBackwardData> checker(handle_cuda());
     NormalRNG default_rng;
     checker.set_dtype(0, dtype::Float32())
@@ -229,6 +234,7 @@ TEST_F(CUDA, SHAKE_CONVOLUTION_BACKWARD_DATA) {
 }
 
 TEST_F(CUDA, SHAKE_CONVOLUTION_BACKWARD_FILTER) {
+    require_compute_capability(6, 1);
     AccuracyShakeChecker<ConvolutionBackwardFilter> checker(handle_cuda());
     NormalRNG default_rng;
     checker.set_dtype(0, dtype::Float32())

dnn/test/cuda/benchmark.cpp

@@ -11,11 +11,11 @@
 #include "test/cuda/fixture.h"
 
 #include "megdnn/oprs.h"
-#include "src/cuda/utils.h"
 #include "test/common/benchmarker.h"
 #include "test/common/tensor.h"
 #include "test/common/timer.h"
 #include "test/common/workspace_wrapper.h"
+#include "test/cuda/utils.h"
 
 namespace megdnn {
 namespace test {
@@ -23,11 +23,7 @@ namespace test {
 #if MEGDNN_WITH_BENCHMARK
 
 TEST_F(CUDA, BENCHMARK_CONVOLUTION_8X8X32) {
-    if (!cuda::is_compute_capability_required(6, 1)) {
-        printf("Skip CUDA.BENCHMARK_CONVOLUTION_8X8X32 test as current device"
-               "doesn't support\n");
-        return;
-    }
+    require_compute_capability(6, 1);
     using Param = param::Convolution;
     auto run_1x1 = [&](size_t N, size_t OC, size_t IC, size_t H, size_t W) {
         Benchmarker<Convolution> benchmarker(handle_cuda());

dnn/test/cuda/convolution.cpp

@@ -13,7 +13,6 @@
 #include "megdnn/dtype.h"
 #include "megdnn/opr_param_defs.h"
 #include "megdnn/oprs.h"
-#include "src/cuda/utils.h"
 #include "test/common/accuracy_shake_checker.h"
 #include "test/common/checker.h"
 #include "test/common/rng.h"
@@ -21,6 +20,7 @@
 #include "test/common/workspace_wrapper.h"
 #include "test/cuda/benchmark.h"
 #include "test/cuda/fixture.h"
+#include "test/cuda/utils.h"
 
 #define V1(x) #x
 #define V(x)  V1(x)
@@ -31,11 +31,7 @@ namespace megdnn {
 namespace test {
 
 TEST_F(CUDA, CONVOLUTION_8X8X32) {
-    if (!cuda::is_compute_capability_required(6, 1)) {
-        printf("Skip CUDA.CONVOLUTION_8X8X32 test as current device"
-               "doesn't support\n");
-        return;
-    }
+    require_compute_capability(6, 1);
 
     using namespace convolution;
     std::vector<TestArg> args;
@@ -116,8 +112,7 @@ TEST_F(CUDA, CONVOLUTION_FORWARD) {
 }
 
 TEST_F(CUDA, CONV_FORWARD_MATMUL_NCHW4) {
-    if (!cuda::is_compute_capability_required(6, 1))
-        return;
+    require_compute_capability(6, 1);
     using namespace convolution;
     Checker<Convolution> checker(handle_cuda());
     UniformIntRNG int_rng{-127, 127};
@@ -205,7 +200,7 @@ TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA) {
                 .set_epsilon(1e-3)
                 .set_param(arg.param)
                 .exec(TensorLayoutArray{filter, dst, src});
-        if (!cuda::is_compute_capability_required(6, 0)) {
+        if (!check_compute_capability(6, 0)) {
             src.dtype = dst.dtype = filter.dtype = dtype::Float16();
             checker.set_rng(0, &rng)
                     .set_rng(1, &rng)
@@ -315,8 +310,7 @@ TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_NHWC) {
 }
 
 TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_CUDNN) {
-    if (cuda::is_compute_capability_required(7, 0))
-        return;
+    require_compute_capability(7, 0);
     using namespace convolution;
     Checker<ConvolutionBackwardData> checker(handle_cuda());
     checker.set_before_exec_callback(
@@ -372,11 +366,7 @@ TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_MATMUL) {
 }
 
 TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_INT8_NCHW4_DP4A) {
-    if (!cuda::is_compute_capability_required(6, 1)) {
-        printf("Skip CUDA.CONVOLUTION_BACKWARD_DATA_INT8_NCHW4_DP4A test as "
-               "current device doesn't support\n");
-        return;
-    }
+    require_compute_capability(6, 1);
 
     using namespace convolution;
     std::vector<TestArg> args = get_args_int8_nchw4_conv_bwd_data();
@@ -430,12 +420,7 @@ TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_INT8_NCHW4_DP4A) {
 }
 
 TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_INT8_NCHW_DP4A) {
-    if (!cuda::is_compute_capability_required(6, 1)) {
-        printf("Skip CUDA.CONVOLUTION_BACKWARD_DATA_INT8_NCHW_DP4A test as "
-               "current device doesn't support\n");
-        return;
-    }
-
+    require_compute_capability(6, 1);
     using namespace convolution;
     std::vector<TestArg> args = get_args_int8_nchw_conv_bwd_data();
     Checker<ConvolutionBackwardData> checker(handle_cuda());
@@ -463,11 +448,7 @@ TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_INT8_NCHW_DP4A) {
 
 #if CUDA_VERSION >= 10020
 TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_INT8_NHWC_IMMA) {
-    if (!cuda::is_compute_capability_required(7, 5)) {
-        printf("Skip CUDA.CONVOLUTION_BACKWARD_DATA_INT8_NHWC_IMMA test as "
-               "current device doesn't support\n");
-        return;
-    }
+    require_compute_capability(7, 5);
 
     using namespace convolution;
     std::vector<TestArg> args = get_args_int8_nhwc_conv_bwd_data();
@@ -527,8 +508,7 @@ TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_INT8_NHWC_IMMA) {
 TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_FAILED_CUDNN7_5) {
     // BRAIN-481 failed on architectures 7.0, remove the following if statement,
     // when cudnn fixed the problem.
-    if (cuda::is_compute_capability_required(7, 0))
-        return;
+    require_compute_capability(7, 0);
     using namespace convolution;
     std::vector<TestArg> args = get_args_cudnn_7_5_failures();
     Checker<ConvolutionBackwardData> checker(handle_cuda());
@@ -662,8 +642,7 @@ TEST_F(CUDA, CONVOLUTION_BACKWARD_FILTER_MATMUL) {
 }
 
 TEST_F(CUDA, CONVOLUTION_BACKWARD_FILTER_CUDNN) {
-    if (cuda::is_compute_capability_required(7, 0))
-        return;
+    require_compute_capability(7, 0);
     using namespace convolution;
     Checker<ConvolutionBackwardFilter> checker(handle_cuda());
     checker.set_before_exec_callback(
@@ -697,8 +676,7 @@ TEST_F(CUDA, CONV_CONFIG_COMBINATIONS) {
 }
 
 TEST_F(CUDA, CONVOLUTION_BACKWARD_DATA_1) {
-    if (cuda::is_compute_capability_required(7, 0))
-        return;
+    require_compute_capability(7, 0);
     using namespace convolution;
     Checker<ConvolutionBackwardData> checker(handle_cuda());
     checker.set_before_exec_callback(AlgoChecker<ConvolutionBackwardData>(

dnn/test/cuda/convolution3d.cpp

@@ -11,20 +11,20 @@
 #include "test/common/convolution3d.h"
 #include "megdnn/opr_param_defs.h"
 #include "megdnn/oprs.h"
-#include "src/cuda/utils.h"
 #include "test/common/benchmarker.h"
 #include "test/common/checker.h"
 #include "test/common/rng.h"
 #include "test/common/tensor.h"
 #include "test/common/workspace_wrapper.h"
 #include "test/cuda/fixture.h"
+#include "test/cuda/utils.h"
 
 namespace megdnn {
 namespace test {
 
 #if 0
 TEST_F(CUDA, CONVOLUTION3D_8X8X32) {
-    if (!cuda::is_compute_capability_required(6, 1)) {
+    if (!check_compute_capability(6, 1)) {
         printf("Skip CUDA.CONVOLUTION_8X8X32 test as current device"
                "doesn't support\n");
         return;

dnn/test/cuda/group_conv.cpp

@@ -15,13 +15,13 @@
 #include "test/common/convolution.h"
 #include "test/cuda/fixture.h"
 
-#include "src/cuda/utils.h"
+#include "test/cuda/utils.h"
 
 namespace megdnn {
 namespace test {
 
 TEST_F(CUDA, GROUP_CONV_FORWARD) {
-    bool is_int_available = cuda::is_compute_capability_required(6, 1);
+    bool is_int_available = check_compute_capability(6, 1);
     auto run = [&](size_t N, size_t IC, size_t IH, size_t IW, size_t FH, size_t FW,
                    size_t OC, size_t /* OH */, size_t /* OW */, size_t PH, size_t PW,
                    size_t SH, size_t SW, size_t DH, size_t DW, size_t group) {

dnn/test/cuda/group_conv3d.cpp

@@ -15,13 +15,13 @@
 #include "test/common/convolution3d.h"
 #include "test/cuda/fixture.h"
 
-#include "src/cuda/utils.h"
+#include "test/cuda/utils.h"
 
 namespace megdnn {
 namespace test {
 
 TEST_F(CUDA, GROUP_CONVOLUTION3D_FORWARD) {
-    bool is_int_available = cuda::is_compute_capability_required(6, 1);
+    bool is_int_available = check_compute_capability(6, 1);
     static_cast<void>(is_int_available);
     auto run = [&](size_t N, size_t IC, size_t ID, size_t IH, size_t IW, size_t FD,
                    size_t FH, size_t FW, size_t OC, size_t PD, size_t PH, size_t PW,

dnn/test/cuda/matrix_mul.cpp

@@ -15,7 +15,6 @@
 #include "test/common/checker.h"
 #include "test/common/matrix_mul.h"
 
-#include "src/cuda/utils.h"
 #if defined(cuda_check)
 #undef cuda_check
 #endif
@@ -130,10 +129,7 @@ TEST_F(CUDA, PEAK_BENCHMARK_MATRIX_MUL_QUANTIZED4x4x32) {
 #endif
 
 TEST_F(CUDA, MATRIX_MUL_INT8x8x32_WITH_SPETIAL_STRIDES) {
-    if (!cuda::is_compute_capability_required(6, 1)) {
-        printf("Skip CUDA.MATRIX_MUL test as current device doesn't support\n");
-        return;
-    }
+    require_compute_capability(6, 1);
     Checker<MatrixMul> checker(handle_cuda());
     using Param = MatrixMul::Param;
     Param param;
@@ -152,10 +148,7 @@ TEST_F(CUDA, MATRIX_MUL_INT8x8x32_WITH_SPETIAL_STRIDES) {
 }
 
 TEST_F(CUDA, MATRIX_MUL_INT8x8x32_NAIVE) {
-    if (!cuda::is_compute_capability_required(6, 1)) {
-        printf("Skip CUDA.MATRIX_MUL test as current device doesn't support\n");
-        return;
-    }
+    require_compute_capability(6, 1);
 
     using Param = MatrixMul::Param;
     UniformIntRNG rng{-128, 127};
@@ -224,16 +217,12 @@ TEST_F(CUDA, MATRIX_MUL_FLOAT_NAIVE) {
 }
 
 TEST_F(CUDA, MATRIX_MUL) {
-    if (cuda::current_device_prop().major < 6) {
-        printf("Skip CUDA.MATRIX_MUL test as current device doesn't support\n");
-        return;
-    }
     Checker<MatrixMul> checker(handle_cuda());
     using Param = MatrixMul::Param;
 
     size_t m = 12, n = 16, k = 20;
 
-    bool is_int_available = cuda::is_compute_capability_required(6, 1);
+    bool is_int_available = check_compute_capability(6, 1);
     std::vector<DType> dtype_array;
     dtype_array.push_back(dtype::Float32());
     dtype_array.push_back(dtype::Float16());

dnn/test/cuda/reduce.cpp

@@ -41,7 +41,7 @@ TEST_F(CUDA, REDUCE) {
     checker.execs({{2, 16 * 16 + 1, 31}, {}});
     checker.execs({{2, 16 * 16 * 16 + 1, 31}, {}});
     checker.execs({{2, 16 * 16 * 16 * 16 + 1, 31}, {}});
-#if MEGDNN_TEGRA_X1
+#if MEGDNN_TEGRA_X1 || MEGDNN_TEGRA_X2
     checker.execs({{2, 8 * 16 * 16 * 16 * 16 + 1, 31}, {}});
 #else
     checker.execs({{2, 16 * 16 * 16 * 16 * 16 + 1, 31}, {}});

dnn/test/cuda/utils.cpp

@@ -18,6 +18,13 @@ bool check_compute_capability(int major, int minor) {
     cuda_check(cudaGetDevice(&dev));
     cudaDeviceProp prop;
     cuda_check(cudaGetDeviceProperties(&prop, dev));
+
+    //! we just skip sm_62 here, which means jetson tx2
+    //! unless require sm_62 explicitly
+    if (prop.major == 6 && prop.minor == 2) {
+        return prop.major == major && prop.minor == minor;
+    }
+
     return prop.major > major || (prop.major == major && prop.minor >= minor);
 }