| 1 | // Tests the phases generated for a CUDA offloading target for different |
| 2 | // combinations of: |
| 3 | // - Number of gpu architectures; |
| 4 | // - Host/device-only compilation; |
| 5 | // - User-requested final phase - binary or assembly. |
| 6 | |
| 7 | // REQUIRES: clang-driver |
| 8 | // REQUIRES: powerpc-registered-target |
| 9 | // REQUIRES: nvptx-registered-target |
| 10 | // REQUIRES: amdgpu-registered-target |
| 11 | // |
| 12 | // Test single gpu architecture with complete compilation. |
| 13 | // |
| 14 | // Test CUDA NVPTX phases. |
| 15 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 16 | // RUN: --cuda-gpu-arch=sm_30 %s 2>&1 \ |
| 17 | // RUN: | FileCheck -check-prefixes=BIN,BIN_NV %s |
| 18 | // |
| 19 | // Test HIP AMDGPU -fgpu-rdc phases. |
| 20 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 21 | // RUN: --cuda-gpu-arch=gfx803 -fgpu-rdc %s 2>&1 \ |
| 22 | // RUN: | FileCheck -check-prefixes=BIN,BIN_AMD,BIN_AMD_RDC %s |
| 23 | // |
| 24 | // Test HIP AMDGPU -fno-gpu-rdc phases (default). |
| 25 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 26 | // RUN: --cuda-gpu-arch=gfx803 %s 2>&1 \ |
| 27 | // RUN: | FileCheck -check-prefixes=BIN,BIN_AMD,BIN_AMD_NRDC %s |
| 28 | // |
| 29 | // BIN_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (host-[[T]]) |
| 30 | // BIN_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (host-[[T]]) |
| 31 | // BIN-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (host-[[T]]) |
| 32 | // BIN-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (host-[[T]]) |
| 33 | // BIN_NV-DAG: [[P3:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (device-[[T]], [[ARCH:sm_30]]) |
| 34 | // BIN_AMD-DAG: [[P3:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (device-[[T]], [[ARCH:gfx803]]) |
| 35 | // BIN-DAG: [[P4:[0-9]+]]: preprocessor, {[[P3]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH]]) |
| 36 | // BIN-DAG: [[P5:[0-9]+]]: compiler, {[[P4]]}, ir, (device-[[T]], [[ARCH]]) |
| 37 | // BIN_NV-DAG: [[P6:[0-9]+]]: backend, {[[P5]]}, assembler, (device-[[T]], [[ARCH]]) |
| 38 | // BIN_NV-DAG: [[P7:[0-9]+]]: assembler, {[[P6]]}, object, (device-[[T]], [[ARCH]]) |
| 39 | // BIN_NV-DAG: [[P8:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE:nvptx64-nvidia-cuda]]:[[ARCH]])" {[[P7]]}, object |
| 40 | // BIN_NV-DAG: [[P9:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE]]:[[ARCH]])" {[[P6]]}, assembler |
| 41 | // BIN_NV-DAG: [[P10:[0-9]+]]: linker, {[[P8]], [[P9]]}, cuda-fatbin, (device-[[T]]) |
| 42 | // BIN_NV-DAG: [[P11:[0-9]+]]: offload, "host-[[T]] (powerpc64le-ibm-linux-gnu)" {[[P2]]}, "device-[[T]] ([[TRIPLE]])" {[[P10]]}, ir |
| 43 | // BIN_NV-DAG: [[P12:[0-9]+]]: backend, {[[P11]]}, assembler, (host-[[T]]) |
| 44 | // BIN_AMD_RDC-DAG: [[P12:[0-9]+]]: backend, {[[P2]]}, assembler, (host-[[T]]) |
| 45 | // BIN_AMD_NRDC-DAG: [[P6:[0-9]+]]: linker, {[[P5]]}, image, (device-hip, [[ARCH]]) |
| 46 | // BIN_AMD_NRDC-DAG: [[P7:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH]])" {[[P6]]}, image |
| 47 | // BIN_AMD_NRDC-DAG: [[P8:[0-9]+]]: linker, {[[P7]]}, hip-fatbin, (device-hip) |
| 48 | // BIN_AMD_NRDC-DAG: [[P11:[0-9]+]]: offload, "host-hip (powerpc64le-ibm-linux-gnu)" {[[P2]]}, "device-hip (amdgcn-amd-amdhsa)" {[[P8]]}, ir |
| 49 | // BIN_AMD_NRDC-DAG: [[P12:[0-9]+]]: backend, {[[P11]]}, assembler, (host-[[T]]) |
| 50 | // BIN-DAG: [[P13:[0-9]+]]: assembler, {[[P12]]}, object, (host-[[T]]) |
| 51 | // BIN-DAG: [[P14:[0-9]+]]: linker, {[[P13]]}, image, (host-[[T]]) |
| 52 | // BIN_AMD_RDC-DAG: [[P15:[0-9]+]]: linker, {[[P5]]}, image, (device-[[T]], [[ARCH]]) |
| 53 | // BIN_AMD_RDC-DAG: [[P16:[0-9]+]]: offload, "host-[[T]] (powerpc64le-ibm-linux-gnu)" {[[P14]]}, |
| 54 | // BIN_AMD_RDC-DAG-SAME: "device-[[T]] ([[TRIPLE:amdgcn-amd-amdhsa]]:[[ARCH]])" {[[P15]]}, object |
| 55 | |
| 56 | // |
| 57 | // Test single gpu architecture up to the assemble phase. |
| 58 | // |
| 59 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 60 | // RUN: --cuda-gpu-arch=sm_30 %s -S 2>&1 \ |
| 61 | // RUN: | FileCheck -check-prefixes=ASM,ASM_NV %s |
| 62 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 63 | // RUN: --cuda-gpu-arch=gfx803 -fgpu-rdc %s -S 2>&1 \ |
| 64 | // RUN: | FileCheck -check-prefixes=ASM,ASM_AMD %s |
| 65 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 66 | // RUN: --cuda-gpu-arch=gfx803 -fcuda-rdc %s -S 2>&1 \ |
| 67 | // RUN: | FileCheck -check-prefixes=ASM,ASM_AMD %s |
| 68 | // ASM_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (device-[[T]], [[ARCH:sm_30]]) |
| 69 | // ASM_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (device-[[T]], [[ARCH:gfx803]]) |
| 70 | // ASM-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH]]) |
| 71 | // ASM-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (device-[[T]], [[ARCH]]) |
| 72 | // ASM_NV-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (device-[[T]], [[ARCH]]) |
| 73 | // ASM_NV-DAG: [[P4:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE:nvptx64-nvidia-cuda|amdgcn-amd-amdhsa]]:[[ARCH]])" {[[P3]]}, assembler |
| 74 | // ASM-DAG: [[P5:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (host-[[T]]) |
| 75 | // ASM-DAG: [[P6:[0-9]+]]: preprocessor, {[[P5]]}, [[T]]-cpp-output, (host-[[T]]) |
| 76 | // ASM-DAG: [[P7:[0-9]+]]: compiler, {[[P6]]}, ir, (host-[[T]]) |
| 77 | // ASM-DAG: [[P8:[0-9]+]]: backend, {[[P7]]}, assembler, (host-[[T]]) |
| 78 | |
| 79 | // |
| 80 | // Test two gpu architectures with complete compilation. |
| 81 | // |
| 82 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 83 | // RUN: --cuda-gpu-arch=sm_30 --cuda-gpu-arch=sm_35 %s 2>&1 \ |
| 84 | // RUN: | FileCheck -check-prefixes=BIN2,BIN2_NV %s |
| 85 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 86 | // RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 -fgpu-rdc %s 2>&1 \ |
| 87 | // RUN: | FileCheck -check-prefixes=BIN2,BIN2_AMD %s |
| 88 | // BIN2_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (host-[[T]]) |
| 89 | // BIN2_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (host-[[T]]) |
| 90 | // BIN2-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (host-[[T]]) |
| 91 | // BIN2-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (host-[[T]]) |
| 92 | // BIN2-DAG: [[P3:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (device-[[T]], [[ARCH1:sm_30|gfx803]]) |
| 93 | // BIN2-DAG: [[P4:[0-9]+]]: preprocessor, {[[P3]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH1]]) |
| 94 | // BIN2-DAG: [[P5:[0-9]+]]: compiler, {[[P4]]}, ir, (device-[[T]], [[ARCH1]]) |
| 95 | // BIN2_NV-DAG: [[P6:[0-9]+]]: backend, {[[P5]]}, assembler, (device-[[T]], [[ARCH1]]) |
| 96 | // BIN2_NV-DAG: [[P7:[0-9]+]]: assembler, {[[P6]]}, object, (device-[[T]], [[ARCH1]]) |
| 97 | // BIN2_NV-DAG: [[P8:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE:nvptx64-nvidia-cuda]]:[[ARCH1]])" {[[P7]]}, object |
| 98 | // BIN2_NV-DAG: [[P9:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE]]:[[ARCH1]])" {[[P6]]}, assembler |
| 99 | // BIN2-DAG: [[P10:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (device-[[T]], [[ARCH2:sm_35|gfx900]]) |
| 100 | // BIN2-DAG: [[P11:[0-9]+]]: preprocessor, {[[P10]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH2]]) |
| 101 | // BIN2-DAG: [[P12:[0-9]+]]: compiler, {[[P11]]}, ir, (device-[[T]], [[ARCH2]]) |
| 102 | // BIN2_NV-DAG: [[P13:[0-9]+]]: backend, {[[P12]]}, assembler, (device-[[T]], [[ARCH2]]) |
| 103 | // BIN2_NV-DAG: [[P14:[0-9]+]]: assembler, {[[P13]]}, object, (device-[[T]], [[ARCH2]]) |
| 104 | // BIN2_NV-DAG: [[P15:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE]]:[[ARCH2]])" {[[P14]]}, object |
| 105 | // BIN2_NV-DAG: [[P16:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE]]:[[ARCH2]])" {[[P13]]}, assembler |
| 106 | // BIN2_NV-DAG: [[P17:[0-9]+]]: linker, {[[P8]], [[P9]], [[P15]], [[P16]]}, cuda-fatbin, (device-[[T]]) |
| 107 | // BIN2_NV-DAG: [[P18:[0-9]+]]: offload, "host-[[T]] (powerpc64le-ibm-linux-gnu)" {[[P2]]}, "device-[[T]] ([[TRIPLE]])" {[[P17]]}, ir |
| 108 | // BIN2_NV-DAG: [[P19:[0-9]+]]: backend, {[[P18]]}, assembler, (host-[[T]]) |
| 109 | // BIN2_AMD-DAG: [[P19:[0-9]+]]: backend, {[[P2]]}, assembler, (host-[[T]]) |
| 110 | // BIN2-DAG: [[P20:[0-9]+]]: assembler, {[[P19]]}, object, (host-[[T]]) |
| 111 | // BIN2-DAG: [[P21:[0-9]+]]: linker, {[[P20]]}, image, (host-[[T]]) |
| 112 | // BIN2_AMD-DAG: [[P22:[0-9]+]]: linker, {[[P5]]}, image, (device-[[T]], [[ARCH1]]) |
| 113 | // BIN2_AMD-DAG: [[P23:[0-9]+]]: linker, {[[P12]]}, image, (device-[[T]], [[ARCH2]]) |
| 114 | // BIN2_AMD-DAG: [[P24:[0-9]+]]: offload, "host-[[T]] (powerpc64le-ibm-linux-gnu)" {[[P21]]}, |
| 115 | // BIN2_AMD-DAG-SAME: "device-[[T]] ([[TRIPLE:amdgcn-amd-amdhsa]]:[[ARCH1]])" {[[P22]]}, |
| 116 | // BIN2_AMD-DAG-SAME: "device-[[T]] ([[TRIPLE:amdgcn-amd-amdhsa]]:[[ARCH2]])" {[[P23]]}, object |
| 117 | |
| 118 | // |
| 119 | // Test two gpu architecturess up to the assemble phase. |
| 120 | // |
| 121 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 122 | // RUN: --cuda-gpu-arch=sm_30 --cuda-gpu-arch=sm_35 %s -S 2>&1 \ |
| 123 | // RUN: | FileCheck -check-prefixes=ASM2,ASM2_NV %s |
| 124 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 125 | // RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 -fgpu-rdc %s -S 2>&1 \ |
| 126 | // RUN: | FileCheck -check-prefixes=ASM2,ASM2_AMD %s |
| 127 | // ASM2_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (device-[[T]], [[ARCH1:sm_30]]) |
| 128 | // ASM2_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (device-[[T]], [[ARCH1:gfx803]]) |
| 129 | // ASM2-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH1]]) |
| 130 | // ASM2-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (device-[[T]], [[ARCH1]]) |
| 131 | // ASM2_NV-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (device-[[T]], [[ARCH1]]) |
| 132 | // ASM2_NV-DAG: [[P4:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE:nvptx64-nvidia-cuda|amdgcn-amd-amdhsa]]:[[ARCH1]])" {[[P3]]}, assembler |
| 133 | // ASM2-DAG: [[P5:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (device-[[T]], [[ARCH2:sm_35|gfx900]]) |
| 134 | // ASM2-DAG: [[P6:[0-9]+]]: preprocessor, {[[P5]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH2]]) |
| 135 | // ASM2-DAG: [[P7:[0-9]+]]: compiler, {[[P6]]}, ir, (device-[[T]], [[ARCH2]]) |
| 136 | // ASM2_NV-DAG: [[P8:[0-9]+]]: backend, {[[P7]]}, assembler, (device-[[T]], [[ARCH2]]) |
| 137 | // ASM2_NV-DAG: [[P9:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE]]:[[ARCH2]])" {[[P8]]}, assembler |
| 138 | // ASM2-DAG: [[P10:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (host-[[T]]) |
| 139 | // ASM2-DAG: [[P11:[0-9]+]]: preprocessor, {[[P10]]}, [[T]]-cpp-output, (host-[[T]]) |
| 140 | // ASM2-DAG: [[P12:[0-9]+]]: compiler, {[[P11]]}, ir, (host-[[T]]) |
| 141 | // ASM2-DAG: [[P13:[0-9]+]]: backend, {[[P12]]}, assembler, (host-[[T]]) |
| 142 | |
| 143 | // |
| 144 | // Test single gpu architecture with complete compilation in host-only |
| 145 | // compilation mode. |
| 146 | // |
| 147 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 148 | // RUN: --cuda-gpu-arch=sm_30 %s --cuda-host-only 2>&1 \ |
| 149 | // RUN: | FileCheck -check-prefixes=HBIN,HBIN_NV %s |
| 150 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 151 | // RUN: --cuda-gpu-arch=gfx803 %s --cuda-host-only 2>&1 \ |
| 152 | // RUN: | FileCheck -check-prefixes=HBIN,HBIN_AMD %s |
| 153 | // HBIN_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (host-[[T]]) |
| 154 | // HBIN_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (host-[[T]]) |
| 155 | // HBIN-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (host-[[T]]) |
| 156 | // HBIN-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (host-[[T]]) |
| 157 | // HBIN-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (host-[[T]]) |
| 158 | // HBIN-DAG: [[P4:[0-9]+]]: assembler, {[[P3]]}, object, (host-[[T]]) |
| 159 | // HBIN-DAG: [[P5:[0-9]+]]: linker, {[[P4]]}, image, (host-[[T]]) |
| 160 | // HBIN-NOT: device |
| 161 | // |
| 162 | // Test single gpu architecture up to the assemble phase in host-only |
| 163 | // compilation mode. |
| 164 | // |
| 165 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 166 | // RUN: --cuda-gpu-arch=sm_30 %s --cuda-host-only -S 2>&1 \ |
| 167 | // RUN: | FileCheck -check-prefixes=HASM,HASM_NV %s |
| 168 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 169 | // RUN: --cuda-gpu-arch=gfx803 %s --cuda-host-only -S 2>&1 \ |
| 170 | // RUN: | FileCheck -check-prefixes=HASM,HASM_AMD %s |
| 171 | // HASM_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (host-[[T]]) |
| 172 | // HASM_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (host-[[T]]) |
| 173 | // HASM-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (host-[[T]]) |
| 174 | // HASM-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (host-[[T]]) |
| 175 | // HASM-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (host-[[T]]) |
| 176 | // HASM-NOT: device |
| 177 | |
| 178 | // |
| 179 | // Test two gpu architectures with complete compilation in host-only |
| 180 | // compilation mode. |
| 181 | // |
| 182 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 183 | // RUN: --cuda-gpu-arch=sm_30 --cuda-gpu-arch=sm_35 %s --cuda-host-only 2>&1 \ |
| 184 | // RUN: | FileCheck -check-prefixes=HBIN2,HBIN2_NV %s |
| 185 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 186 | // RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s --cuda-host-only 2>&1 \ |
| 187 | // RUN: | FileCheck -check-prefixes=HBIN2,HBIN2_AMD %s |
| 188 | // HBIN2_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (host-[[T]]) |
| 189 | // HBIN2_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (host-[[T]]) |
| 190 | // HBIN2-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (host-[[T]]) |
| 191 | // HBIN2-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (host-[[T]]) |
| 192 | // HBIN2-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (host-[[T]]) |
| 193 | // HBIN2-DAG: [[P4:[0-9]+]]: assembler, {[[P3]]}, object, (host-[[T]]) |
| 194 | // HBIN2-DAG: [[P5:[0-9]+]]: linker, {[[P4]]}, image, (host-[[T]]) |
| 195 | // HBIN2-NOT: device |
| 196 | |
| 197 | // |
| 198 | // Test two gpu architectures up to the assemble phase in host-only |
| 199 | // compilation mode. |
| 200 | // |
| 201 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 202 | // RUN: --cuda-gpu-arch=sm_30 --cuda-gpu-arch=sm_35 %s --cuda-host-only -S \ |
| 203 | // RUN: 2>&1 | FileCheck -check-prefixes=HASM2,HASM2_NV %s |
| 204 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 205 | // RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s --cuda-host-only -S \ |
| 206 | // RUN: 2>&1 | FileCheck -check-prefixes=HASM2,HASM2_AMD %s |
| 207 | // HASM2_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (host-[[T]]) |
| 208 | // HASM2_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (host-[[T]]) |
| 209 | // HASM2-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (host-[[T]]) |
| 210 | // HASM2-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (host-[[T]]) |
| 211 | // HASM2-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (host-[[T]]) |
| 212 | // HASM2-NOT: device |
| 213 | |
| 214 | // |
| 215 | // Test single gpu architecture with complete compilation in device-only |
| 216 | // compilation mode. |
| 217 | // |
| 218 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 219 | // RUN: --cuda-gpu-arch=sm_30 %s --cuda-device-only 2>&1 \ |
| 220 | // RUN: | FileCheck -check-prefixes=DBIN,DBIN_NV %s |
| 221 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 222 | // RUN: --cuda-gpu-arch=gfx803 %s --cuda-device-only 2>&1 \ |
| 223 | // RUN: | FileCheck -check-prefixes=DBIN,DBIN_AMD %s |
| 224 | // DBIN_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (device-[[T]], [[ARCH:sm_30]]) |
| 225 | // DBIN_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (device-[[T]], [[ARCH:gfx803]]) |
| 226 | // DBIN-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH]]) |
| 227 | // DBIN-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (device-[[T]], [[ARCH]]) |
| 228 | // DBIN_NV-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (device-[[T]], [[ARCH]]) |
| 229 | // DBIN_NV-DAG: [[P4:[0-9]+]]: assembler, {[[P3]]}, object, (device-[[T]], [[ARCH]]) |
| 230 | // DBIN_NV-DAG: [[P5:[0-9]+]]: offload, "device-[[T]] (nvptx64-nvidia-cuda:[[ARCH]])" {[[P4]]}, object |
| 231 | // DBIN-NOT: host |
| 232 | // |
| 233 | // Test single gpu architecture up to the assemble phase in device-only |
| 234 | // compilation mode. |
| 235 | // |
| 236 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 237 | // RUN: --cuda-gpu-arch=sm_30 %s --cuda-device-only -S 2>&1 \ |
| 238 | // RUN: | FileCheck -check-prefixes=DASM,DASM_NV %s |
| 239 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 240 | // RUN: --cuda-gpu-arch=gfx803 %s --cuda-device-only -S 2>&1 \ |
| 241 | // RUN: | FileCheck -check-prefixes=DASM,DASM_AMD %s |
| 242 | // DASM_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (device-[[T]], [[ARCH:sm_30]]) |
| 243 | // DASM_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (device-[[T]], [[ARCH:gfx803]]) |
| 244 | // DASM-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH]]) |
| 245 | // DASM-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (device-[[T]], [[ARCH]]) |
| 246 | // DASM_NV-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (device-[[T]], [[ARCH]]) |
| 247 | // DASM_NV-DAG: [[P4:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE:nvptx64-nvidia-cuda|amdgcn-amd-amdhsa]]:[[ARCH]])" {[[P3]]}, assembler |
| 248 | // DASM-NOT: host |
| 249 | |
| 250 | // |
| 251 | // Test two gpu architectures with complete compilation in device-only |
| 252 | // compilation mode. |
| 253 | // |
| 254 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 255 | // RUN: --cuda-gpu-arch=sm_30 --cuda-gpu-arch=sm_35 %s --cuda-device-only 2>&1 \ |
| 256 | // RUN: | FileCheck -check-prefixes=DBIN2,DBIN2_NV %s |
| 257 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 258 | // RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s --cuda-device-only \ |
| 259 | // RUN: 2>&1 | FileCheck -check-prefixes=DBIN2,DBIN2_AMD %s |
| 260 | // DBIN2_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (device-[[T]], [[ARCH:sm_30]]) |
| 261 | // DBIN2_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (device-[[T]], [[ARCH:gfx803]]) |
| 262 | // DBIN2-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH]]) |
| 263 | // DBIN2-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (device-[[T]], [[ARCH]]) |
| 264 | // DBIN2_NV-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (device-[[T]], [[ARCH]]) |
| 265 | // DBIN2_NV-DAG: [[P4:[0-9]+]]: assembler, {[[P3]]}, object, (device-[[T]], [[ARCH]]) |
| 266 | // DBIN2_NV-DAG: [[P5:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE:nvptx64-nvidia-cuda]]:[[ARCH]])" {[[P4]]}, object |
| 267 | // DBIN2-DAG: [[P6:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (device-[[T]], [[ARCH2:sm_35|gfx900]]) |
| 268 | // DBIN2-DAG: [[P7:[0-9]+]]: preprocessor, {[[P6]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH2]]) |
| 269 | // DBIN2-DAG: [[P8:[0-9]+]]: compiler, {[[P7]]}, ir, (device-[[T]], [[ARCH2]]) |
| 270 | // DBIN2_NV-DAG: [[P9:[0-9]+]]: backend, {[[P8]]}, assembler, (device-[[T]], [[ARCH2]]) |
| 271 | // DBIN2_NV-DAG: [[P10:[0-9]+]]: assembler, {[[P9]]}, object, (device-[[T]], [[ARCH2]]) |
| 272 | // DBIN2_NV-DAG: [[P11:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE]]:[[ARCH2]])" {[[P10]]}, object |
| 273 | // DBIN2-NOT: host |
| 274 | // |
| 275 | // Test two gpu architectures up to the assemble phase in device-only |
| 276 | // compilation mode. |
| 277 | // |
| 278 | // RUN: %clang -target powerpc64le-ibm-linux-gnu -ccc-print-phases \ |
| 279 | // RUN: --cuda-gpu-arch=sm_30 --cuda-gpu-arch=sm_35 %s --cuda-device-only -S \ |
| 280 | // RUN: 2>&1 | FileCheck -check-prefixes=DASM2,DASM2_NV %s |
| 281 | // RUN: %clang -x hip -target powerpc64le-ibm-linux-gnu \ |
| 282 | // RUN: -ccc-print-phases --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \ |
| 283 | // RUN: --cuda-device-only -S 2>&1 \ |
| 284 | // RUN: | FileCheck -check-prefixes=DASM2,DASM2_AMD %s |
| 285 | // DASM2_NV-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:cuda]], (device-[[T]], [[ARCH:sm_30]]) |
| 286 | // DASM2_AMD-DAG: [[P0:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T:hip]], (device-[[T]], [[ARCH:gfx803]]) |
| 287 | // DASM2-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH]]) |
| 288 | // DASM2-DAG: [[P2:[0-9]+]]: compiler, {[[P1]]}, ir, (device-[[T]], [[ARCH]]) |
| 289 | // DASM2_NV-DAG: [[P3:[0-9]+]]: backend, {[[P2]]}, assembler, (device-[[T]], [[ARCH]]) |
| 290 | // DASM2_NV-DAG: [[P4:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE:nvptx64-nvidia-cuda|amdgcn-amd-amdhsa]]:[[ARCH]])" {[[P3]]}, assembler |
| 291 | // DASM2-DAG: [[P5:[0-9]+]]: input, "{{.*}}cuda-phases.cu", [[T]], (device-[[T]], [[ARCH2:sm_35|gfx900]]) |
| 292 | // DASM2-DAG: [[P6:[0-9]+]]: preprocessor, {[[P5]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH2]]) |
| 293 | // DASM2-DAG: [[P7:[0-9]+]]: compiler, {[[P6]]}, ir, (device-[[T]], [[ARCH2]]) |
| 294 | // DASM2_NV-DAG: [[P8:[0-9]+]]: backend, {[[P7]]}, assembler, (device-[[T]], [[ARCH2]]) |
| 295 | // DASM2_NV-DAG: [[P9:[0-9]+]]: offload, "device-[[T]] ([[TRIPLE]]:[[ARCH2]])" {[[P8]]}, assembler |
| 296 | // DASM2-NOT: host |
| 297 | |