XFAIL: cuda tracker - Githubissues

steffenlarsen commented 3 years ago

This issue is for tracking XFAIL:cuda tests. This is a continuation of https://github.com/intel/llvm/issues/1542.

The following test currently have XFAIL:cuda in them: SYCL/Basic/reqd_work_group_size.cpp (Enabled in https://github.com/intel/llvm-test-suite/pull/278) SYCL/Config/select_device.cpp SYCL/DeviceLib/built-ins/nan.cpp (Disabled https://github.com/intel/llvm-test-suite/pull/336) SYCL/HierPar/hier_par_wgscope_O0.cpp (Disabled in https://github.com/intel/llvm-test-suite/pull/290) SYCL/KernelAndProgram/basic-program.cpp SYCL/KernelAndProgram/build-log.cpp SYCL/KernelAndProgram/cache-build-result.cpp SYCL/KernelAndProgram/get-options.cpp SYCL/KernelAndProgram/kernel-and-program.cpp SYCL/Regression/atomic_load.cpp (Enabled in https://github.com/intel/llvm-test-suite/pull/339) ~~SYCL/Regression/implicit_offset_debug_info.cpp~~ (Enabled with https://github.com/intel/llvm-test-suite/pull/264) SYCL/Sampler/normalized-clamp-linear-float.cpp SYCL/Sampler/normalized-clampedge-linear-float.cpp SYCL/Sampler/normalized-mirror-linear-float.cpp SYCL/Sampler/normalized-mirror-nearest.cpp SYCL/Sampler/normalized-none-linear-float.cpp SYCL/Sampler/normalized-repeat-linear-float.cpp SYCL/Sampler/normalized-repeat-nearest.cpp SYCL/Sampler/unnormalized-clamp-linear-float.cpp SYCL/Sampler/unnormalized-clampedge-linear-float.cpp SYCL/Sampler/unnormalized-none-linear-float.cpp SYCL/Scheduler/CommandCleanupThreadSafety.cpp SYCL/Scheduler/MemObjRemapping.cpp (Enabled in https://github.com/intel/llvm-test-suite/pull/339) SYCL/USM/prefetch.cpp

vmaksimo commented 3 years ago

SYCL/HierPar/hier_par_wgscope_O0.cpp (disabled in https://github.com/intel/llvm-test-suite/pull/290)

t4c1 commented 2 years ago

Here is an updated list of tests with XFAIL: cuda:

AtomicRef/*_generic* (PR opened: https://github.com/intel/llvm/pull/5849, but blocked by the remangler) Config/select_device.cpp Regression/kernel_bundle_ignore_sycl_external.cpp Sampler/normalized-mirror-linear-float.cpp Sampler/unnormalized-clamp-linear-float.cpp Sampler/unnormalized-clampedge-linear-float.cpp Sampler/unnormalized-none-linear-float.cpp Sampler/normalized-clamp-linear-float.cpp Sampler/normalized-repeat-linear-float.cpp Sampler/normalized-none-linear-float.cpp Sampler/normalized-clampedge-linear-float.cpp Sampler/normalized-repeat-nearest.cpp Sampler/normalized-mirror-nearest.cpp KernelAndProgram/undefined-symbol.cpp KernelAndProgram/cache-build-result.cpp KernelAndProgram/build-log.cpp Printf/int.cpp Scheduler/CommandCleanupThreadSafety.cpp

Here is also a list of tests with UNSUPPORTED: cuda (ignoring ESIMD and DeprecatedFeatures folders) that in fact pass:

SpecConstants/2020/host_apis.cpp SpecConstants/2020/handler-api.cpp SpecConstants/2020/vector-convolution-demo.cpp SpecConstants/2020/non_native/cuda.cpp SpecConstants/2020/kernel-bundle-api.cpp GroupAlgorithm/reduce_sycl2020.cpp GroupAlgorithm/inclusive_scan_sycl2020.cpp GroupAlgorithm/exclusive_scan_sycl2020.cpp GroupAlgorithm/exclusive_scan_over_group.cpp Assert/assert_in_kernels_ndebug.cpp DeviceLib/ITTAnnotations/atomic.cpp DeviceLib/ITTAnnotations/barrier.cpp Regression/image_access.cpp Basic/image/image_write.cpp Basic/image/image_sample.cpp Basic/image/image_array.cpp

Here is also a list of tests with UNSUPPORTED: cuda (ignoring ESIMD and DeprecatedFeatures folders) that seem to need some work:

Assert/assert_in_kernels.cpp Assert/assert_in_multiple_tus.cpp Assert/assert_in_multiple_tus_one_ndebug.cpp Assert/assert_in_one_kernel.cpp Assert/assert_in_simultaneous_kernels.cpp Assert/assert_in_simultaneously_multiple_tus.cpp Assert/assert_in_simultaneously_multiple_tus_one_ndebug.cpp BFloat16/bfloat16_type.cpp Basic/image/image_accessor_readsampler.cpp Basic/image/image_accessor_readwrite.cpp Basic/image/image_accessor_readwrite_half.cpp Basic/image/image_max_size.cpp Basic/image/srgba-read.cpp DeviceLib/built-ins/printf.cpp DiscardEvents/discard_events_using_assert.cpp KernelAndProgram/kernel-bundle-merge-options-env.cpp KernelAndProgram/kernel-bundle-merge-options.cpp KernelAndProgram/multiple-kernel-linking.cpp Plugin/sycl-ls-gpu-default-any.cpp Regression/commandlist/gpu.cpp Regression/device_num.cpp Sampler/unnormalized-clamp-nearest.cpp Sampler/unnormalized-clampedge-nearest.cpp Sampler/unnormalized-none-nearest.cpp Scheduler/HostAccDestruction.cpp SeparateCompile/test.cpp SubGroup/reduce_spirv13.cpp SubGroup/reduce_spirv13_fp16.cpp SubGroup/scan_spirv13.cpp SubGroup/scan_spirv13_fp16.cpp

AerialMantis commented 2 years ago

Here is a list of the pull requests for addressing issues in the UNSUPPORTED: cuda tests which can be fixed:

Regression/image_access.cpp, Basic/image/image_sample.cpp, Basic/image/image_array.cpp -> https://github.com/intel/llvm-test-suite/pull/1131
Sampler/unnormalized-clamp-nearest.cpp, Sampler/unnormalized-clampedge-nearest.cpp, Sampler/unnormalized-none-nearest.cpp ->https://github.com/intel/llvm/pull/6335 and https://github.com/intel/llvm-test-suite/pull/1062
Config/select_device.cpp -> https://github.com/intel/llvm-test-suite/pull/1086
Scheduler/CommandCleanupThreadSafety.cpp -> https://github.com/intel/llvm-test-suite/pull/1087
DeviceLib/built-ins/printf.cpp -> https://github.com/intel/llvm-test-suite/pull/1096
Regression/commandlist/gpu.cpp -> https://github.com/intel/llvm-test-suite/pull/1088
Regression/device_num.cpp -> https://github.com/intel/llvm-test-suite/pull/1081
Scheduler/HostAccDestruction.cpp -> https://github.com/intel/llvm-test-suite/pull/1085
Assert/assert_in_kernels.cpp, Assert/assert_in_multiple_tus.cpp, Assert/assert_in_multiple_tus_one_ndebug.cpp, Assert/assert_in_one_kernel.cpp, Assert/assert_in_simultaneous_kernels.cpp, Assert/assert_in_simultaneously_multiple_tus.cpp, Assert/assert_in_simultaneously_multiple_tus_one_ndebug.cpp -> https://github.com/intel/llvm/pull/6370 and https://github.com/intel/llvm-test-suite/pull/1070
KernelAndProgram/kernel-bundle-merge-options.cpp, KernelAndProgram/kernel-bundle-merge-options-env.cpp, KernelAndProgram/multiple-kernel-linking.cpp -> https://github.com/intel/llvm-test-suite/pull/1093
SubGroup/reduce_spirv13.cpp, SubGroup/scan_spirv13.cpp -> https://github.com/intel/llvm/pull/6405 and https://github.com/intel/llvm-test-suite/pull/1077

The Printf/int.cpp was not fixed as it is failing due to an Nvidia bug, which has been reported here - https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#format-specifiers

After some investigation we also found that the following tests are not expected to pass with the CUDA backend:

BFloat16/bfloat16_type.cpp
SeparateCompile/test.cpp
KernelAndProgram/cache-build-result.cpp
KernelAndProgram/build-log.cpp
KernelAndProgram/undefined-symbol.cpp

pvchupin commented 2 years ago

Here is a list of the pull requests ...

@AerialMantis, feel free to push individual PRs. It seems some a stuck for no apparent reason...

jle-quel commented 2 years ago

Here is an updated list of the tests marked as XFAIL and UNSUPPORTED for CUDA:

[+] XFAIL SYCL/AtomicRef/load_generic_local.cpp SYCL/AtomicRef/store_generic_local.cpp SYCL/AtomicRef/add_generic_native_fp.cpp SYCL/AtomicRef/or_generic_local.cpp SYCL/AtomicRef/add_generic.cpp SYCL/AtomicRef/add_generic_local.cpp SYCL/AtomicRef/add_generic_local_native_fp.cpp SYCL/AtomicRef/max_generic.cpp SYCL/AtomicRef/max_generic_local.cpp SYCL/AtomicRef/max_generic_local_native_fp.cpp SYCL/AtomicRef/max_generic_native_fp.cpp SYCL/AtomicRef/exchange_generic_local.cpp SYCL/AtomicRef/sub_generic_native_fp.cpp SYCL/AtomicRef/sub_generic_local_native_fp.cpp SYCL/AtomicRef/and_generic_local.cpp SYCL/AtomicRef/and_generic.cpp SYCL/AtomicRef/sub_generic_local.cpp SYCL/AtomicRef/assignment_atomic64_generic.cpp SYCL/AtomicRef/or_generic.cpp SYCL/AtomicRef/compare_exchange_generic_local.cpp SYCL/AtomicRef/store_generic.cpp SYCL/AtomicRef/compare_exchange_generic.cpp SYCL/AtomicRef/load_generic.cpp SYCL/AtomicRef/min_generic.cpp SYCL/AtomicRef/min_generic_local.cpp SYCL/AtomicRef/min_generic_native_fp.cpp SYCL/AtomicRef/min_generic_local_native_fp.cpp SYCL/AtomicRef/xor_generic.cpp SYCL/AtomicRef/sub_generic.cpp SYCL/AtomicRef/exchange_generic.cpp SYCL/AtomicRef/assignment_generic.cpp SYCL/AtomicRef/xor_generic_local.cpp SYCL/Regression/kernel_bundle_ignore_sycl_external.cpp SYCL/Printf/int.cpp SYCL/Sampler/normalized-clampedge-linear-float.cpp SYCL/Sampler/normalized-mirror-linear-float.cpp SYCL/Sampler/unnormalized-clampedge-linear-float.cpp SYCL/Sampler/normalized-repeat-linear-float.cpp SYCL/Sampler/unnormalized-none-linear-float.cpp SYCL/Sampler/normalized-mirror-nearest.cpp SYCL/Sampler/normalized-clamp-linear-float.cpp SYCL/Sampler/normalized-none-linear-float.cpp SYCL/Sampler/unnormalized-clamp-linear-float.cpp SYCL/Sampler/normalized-repeat-nearest.cpp

[+] UNSUPPORTED SYCL/DeviceLib/imf_simd_emulate_test.cpp SYCL/DeviceLib/imf_fp64_test.cpp SYCL/DeviceLib/ITTAnnotations/atomic.cpp SYCL/DeviceLib/ITTAnnotations/barrier.cpp SYCL/DeviceLib/imf_fp32_test.cpp SYCL/DeprecatedFeatures/deprecated_intel_ext_device.cpp SYCL/Plugin/sycl-ls-gpu-default-any.cpp SYCL/Plugin/sycl-ls.cpp SYCL/DiscardEvents/discard_events_using_assert.cpp SYCL/InlineAsm/asm_float_neg.cpp SYCL/InlineAsm/asm_mul.cpp SYCL/InlineAsm/asm_no_operands.cpp SYCL/InlineAsm/Negative/asm_wrong_declare.cpp SYCL/InlineAsm/Negative/asm_illegal_exec_size.cpp SYCL/InlineAsm/Negative/asm_bad_opcode.cpp SYCL/InlineAsm/Negative/asm_bad_operand_syntax.cpp SYCL/InlineAsm/Negative/asm_duplicate_label.cpp SYCL/InlineAsm/Negative/asm_undefined_decl.cpp SYCL/InlineAsm/Negative/asm_undefined_pred.cpp SYCL/InlineAsm/Negative/asm_missing_label.cpp SYCL/InlineAsm/Negative/asm_missing_region.cpp SYCL/InlineAsm/Negative/asm_simple.cpp SYCL/InlineAsm/malloc_shared_in_out_dif.cpp SYCL/InlineAsm/asm_no_output.cpp SYCL/InlineAsm/asm_16_no_opts.cpp SYCL/InlineAsm/asm_16_matrix_mult.cpp SYCL/InlineAsm/asm_16_empty.cpp SYCL/InlineAsm/asm_if.cpp SYCL/InlineAsm/malloc_shared_no_input.cpp SYCL/InlineAsm/asm_multiple_instructions.cpp SYCL/InlineAsm/asm_decl_in_scope.cpp SYCL/InlineAsm/asm_16_no_input_int.cpp SYCL/InlineAsm/asm_plus_mod.cpp SYCL/InlineAsm/asm_arbitrary_ops_order.cpp SYCL/InlineAsm/malloc_shared_32.cpp SYCL/InlineAsm/asm_loop.cpp SYCL/InlineAsm/asm_8_empty.cpp SYCL/InlineAsm/asm_switch.cpp SYCL/InlineAsm/letter_example.cpp SYCL/InlineAsm/asm_float_add.cpp SYCL/InlineAsm/asm_8_no_input_int.cpp SYCL/InlineAsm/asm_float_imm_arg.cpp SYCL/InlineAsm/asm_imm_arg.cpp SYCL/Assert/assert_in_simultaneously_multiple_tus_one_ndebug.cpp SYCL/Assert/assert_in_simultaneous_kernels.cpp SYCL/Assert/assert_in_simultaneously_multiple_tus.cpp SYCL/SpecConstants/2020/non_native/gpu.cpp SYCL/SpecConstants/2020/non_native/multiple-targets.cpp SYCL/USM/dep_events.cpp SYCL/USM/prefetch.cpp SYCL/SeparateCompile/test.cpp SYCL/BFloat16/bfloat16_type.cpp SYCL/AOT/gpu.cpp SYCL/AOT/multiple-devices.cpp SYCL/ESIMD/acc_gather_scatter_rgba.cpp SYCL/ESIMD/regression/sycl_esimd_mixed_unnamed.cpp SYCL/ESIMD/regression/usm_gather_scatter_32.cpp SYCL/ESIMD/regression/big_const_initializer.cpp SYCL/ESIMD/regression/unused_load.cpp SYCL/ESIMD/regression/operator_decrement.cpp SYCL/ESIMD/regression/abs_fix_test.cpp SYCL/ESIMD/regression/dgetrf_ref.cpp SYCL/ESIMD/regression/copyto_char_test.cpp SYCL/ESIMD/regression/replicate_bug.cpp SYCL/ESIMD/regression/variable_gather_mask.cpp SYCL/ESIMD/regression/rol_ror_compilation_test.cpp SYCL/ESIMD/regression/globals.cpp SYCL/ESIMD/regression/fmod_compatibility_test.cpp SYCL/ESIMD/regression/complex-lib-lin.cpp SYCL/ESIMD/regression/double_conversion.cpp SYCL/ESIMD/regression/half_conversion_test.cpp SYCL/ESIMD/regression/dgetrf.cpp SYCL/ESIMD/regression/store_zero_const.cpp SYCL/ESIMD/regression/noinline_bypointers_vadd.cpp SYCL/ESIMD/regression/dgetrf_8x8.cpp SYCL/ESIMD/regression/bit_shift_compilation_test.cpp SYCL/ESIMD/stencil2.cpp SYCL/ESIMD/noinline_call_from_func.cpp SYCL/ESIMD/vadd_usm.cpp SYCL/ESIMD/preemption.cpp SYCL/ESIMD/fp_call_recursive.cpp SYCL/ESIMD/api/svm_gather_scatter.cpp SYCL/ESIMD/api/esimd_merge.cpp SYCL/ESIMD/api/simd_binop_integer_promotion.cpp SYCL/ESIMD/api/simd_any_all.cpp SYCL/ESIMD/api/bin_and_cmp_ops_heavy.cpp SYCL/ESIMD/api/replicate_smoke.cpp SYCL/ESIMD/api/slm_gather_scatter_heavy.cpp SYCL/ESIMD/api/functional/operators/operator_decrement_and_increment_accuracy_fp_extra.cpp SYCL/ESIMD/api/functional/operators/operator_bitwise_not_uint.cpp SYCL/ESIMD/api/functional/operators/operator_decrement_and_increment_core.cpp SYCL/ESIMD/api/functional/operators/operator_decrement_and_increment_fp_extra.cpp SYCL/ESIMD/api/functional/operators/operator_bitwise_not_sint.cpp SYCL/ESIMD/api/functional/operators/operator_decrement_and_increment_accuracy_core.cpp SYCL/ESIMD/api/functional/operators/operator_logical_not.cpp SYCL/ESIMD/api/functional/ctors/ctor_converting_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_broadcast_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_vector_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_fill_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_load_usm_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_default_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_broadcast_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_vector_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_array_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_fill_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_load_acc_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_copy_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_copy_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_fill_accuracy_fp.cpp SYCL/ESIMD/api/functional/ctors/ctor_move_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_array_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_move_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_fill_accuracy_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_load_acc_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_load_usm_core.cpp SYCL/ESIMD/api/functional/ctors/ctor_converting_fp_extra.cpp SYCL/ESIMD/api/functional/ctors/ctor_default_fp_extra.cpp SYCL/ESIMD/api/functional/functions/functions_select_simd_view_rval_fp_extra.cpp SYCL/ESIMD/api/functional/functions/functions_select_2d_core.cpp SYCL/ESIMD/api/functional/functions/functions_select_rvalue_fp_extra.cpp SYCL/ESIMD/api/functional/functions/functions_select_lvalue_fp_extra.cpp SYCL/ESIMD/api/functional/functions/functions_select_simd_view_rval_core.cpp SYCL/ESIMD/api/functional/functions/functions_select_rvalue_core.cpp SYCL/ESIMD/api/functional/functions/functions_select_2d_fp_extra.cpp SYCL/ESIMD/api/functional/functions/functions_select_lvalue_core.cpp SYCL/ESIMD/api/simd_memory_access.cpp SYCL/ESIMD/api/slm_gather_scatter_rgba.cpp SYCL/ESIMD/api/simd_view_select_2d_int.cpp SYCL/ESIMD/api/simd_view_subscript_operator.cpp SYCL/ESIMD/api/simd_subscript_operator.cpp SYCL/ESIMD/api/esimd_bit_ops.cpp SYCL/ESIMD/api/esimd_pack_unpack_mask.cpp SYCL/ESIMD/api/slm_gather_scatter.cpp SYCL/ESIMD/api/simd_view_copy_move_assign.cpp SYCL/ESIMD/api/saturation_smoke.cpp SYCL/ESIMD/api/simd_view_negation_operator.cpp SYCL/ESIMD/api/simd_view_select_2d_fp.cpp SYCL/ESIMD/api/simd_mask.cpp SYCL/ESIMD/api/esimd_rgba_smoke.cpp SYCL/ESIMD/api/ballot.cpp SYCL/ESIMD/api/simd_copy_to_from.cpp SYCL/ESIMD/api/unary_ops_heavy.cpp SYCL/ESIMD/api/simd_negation_operator.cpp SYCL/ESIMD/aot_mixed.cpp SYCL/ESIMD/radix_sort.cpp SYCL/ESIMD/dpas/dpas_tf32.cpp SYCL/ESIMD/dpas/dpas_test3.cpp SYCL/ESIMD/dpas/dpasw_test.cpp SYCL/ESIMD/dpas/dpas_test2.cpp SYCL/ESIMD/dpas/dpas_test1.cpp SYCL/ESIMD/dpas/dpas_bf16.cpp SYCL/ESIMD/dpas/dpas_fp16.cpp SYCL/ESIMD/dpas/dpas_int.cpp SYCL/ESIMD/accessor_stateless.cpp SYCL/ESIMD/matrix_transpose.cpp SYCL/ESIMD/esimd_check_vc_codegen.cpp SYCL/ESIMD/BitonicSortK.cpp SYCL/ESIMD/accessor.cpp SYCL/ESIMD/vec_arg_call_conv_smoke.cpp SYCL/ESIMD/vadd_2d.cpp SYCL/ESIMD/accessor_gather_scatter_stateless.cpp SYCL/ESIMD/vadd_1d.cpp SYCL/ESIMD/fp_call_from_func.cpp SYCL/ESIMD/ext_math.cpp SYCL/ESIMD/noinline_args_size/noinline_args_char_int_size_96.cpp SYCL/ESIMD/noinline_args_size/noinline_args_char_int_size_64.cpp SYCL/ESIMD/noinline_args_size/noinline_args_int_size_512.cpp SYCL/ESIMD/noinline_args_size/noinline_args_char_int_size_192.cpp SYCL/ESIMD/noinline_args_size/noinline_args_int_size_96.cpp SYCL/ESIMD/noinline_args_size/noinline_args_int_size_192.cpp SYCL/ESIMD/noinline_args_size/noinline_args_char_int_size_256.cpp SYCL/ESIMD/noinline_args_size/noinline_args_int_size_256.cpp SYCL/ESIMD/noinline_args_size/noinline_args_char_int_size_512.cpp SYCL/ESIMD/noinline_args_size/noinline_args_int_size_64.cpp SYCL/ESIMD/vadd_raw_send.cpp SYCL/ESIMD/mandelbrot/mandelbrot_spec.cpp SYCL/ESIMD/mandelbrot/mandelbrot.cpp SYCL/ESIMD/fp_in_phi.cpp SYCL/ESIMD/Prefix_Local_sum2.cpp SYCL/ESIMD/matrix_transpose2.cpp SYCL/ESIMD/Prefix_Local_sum1.cpp SYCL/ESIMD/spec_const/spec_const_uint.cpp SYCL/ESIMD/spec_const/spec_const_char.cpp SYCL/ESIMD/spec_const/spec_const_int64.cpp SYCL/ESIMD/spec_const/spec_const_float.cpp SYCL/ESIMD/spec_const/spec_const_bool.cpp SYCL/ESIMD/spec_const/spec_const_ushort.cpp SYCL/ESIMD/spec_const/spec_const_short.cpp SYCL/ESIMD/spec_const/spec_const_double.cpp SYCL/ESIMD/spec_const/unused_spec_const.cpp SYCL/ESIMD/spec_const/spec_const_int.cpp SYCL/ESIMD/spec_const/spec_const_uint64.cpp SYCL/ESIMD/spec_const/spec_const_uchar.cpp SYCL/ESIMD/fp_args_size/fp_args_char_int_size_96.cpp SYCL/ESIMD/fp_args_size/fp_args_char_int_size_256.cpp SYCL/ESIMD/fp_args_size/fp_args_int_size_256.cpp SYCL/ESIMD/fp_args_size/fp_args_int_size_512.cpp SYCL/ESIMD/fp_args_size/fp_args_char_int_size_192.cpp SYCL/ESIMD/fp_args_size/fp_args_int_size_64.cpp SYCL/ESIMD/fp_args_size/fp_args_int_size_96.cpp SYCL/ESIMD/fp_args_size/fp_args_char_int_size_64.cpp SYCL/ESIMD/fp_args_size/fp_args_char_int_size_512.cpp SYCL/ESIMD/fp_args_size/fp_args_int_size_192.cpp SYCL/ESIMD/printf.cpp SYCL/ESIMD/matrix_transpose_glb.cpp SYCL/ESIMD/usm_gather_scatter_rgba.cpp SYCL/ESIMD/fp_in_select.cpp SYCL/ESIMD/accessor_load_store_stateless.cpp SYCL/ESIMD/dword_atomic_cmpxchg.cpp SYCL/ESIMD/histogram_raw_send.cpp SYCL/ESIMD/histogram_256_slm_spec_2020.cpp SYCL/ESIMD/accessor_gather_scatter.cpp SYCL/ESIMD/Prefix_Local_sum3.cpp SYCL/ESIMD/slm_barrier.cpp SYCL/ESIMD/vec_arg_call_conv_ext.cpp SYCL/ESIMD/reduction.cpp SYCL/ESIMD/lsc/lsc_slm_scatter.cpp SYCL/ESIMD/lsc/lsc_surf_load_u16u32.cpp SYCL/ESIMD/lsc/lsc_block_store_u32.cpp SYCL/ESIMD/lsc/lsc_surf_store_u8_u16.cpp SYCL/ESIMD/lsc/lsc_block_store_u64.cpp SYCL/ESIMD/lsc/lsc_surf_prefetch_u64.cpp SYCL/ESIMD/lsc/lsc_block_load_u64.cpp SYCL/ESIMD/lsc/lsc_block_prefetch_u64.cpp SYCL/ESIMD/lsc/lsc_usm_store_u8_u16.cpp SYCL/ESIMD/lsc/lsc_usm_store_u8u32.cpp SYCL/ESIMD/lsc/lsc_surf.cpp SYCL/ESIMD/lsc/lsc_surf_load_u64.cpp SYCL/ESIMD/lsc/lsc_surf_stateless.cpp SYCL/ESIMD/lsc/lsc_block_load_u8.cpp SYCL/ESIMD/lsc/lsc_surf_prefetch_u16u32.cpp SYCL/ESIMD/lsc/lsc_usm_prefetch_u64.cpp SYCL/ESIMD/lsc/lsc_usm_atomic_cachehint.cpp SYCL/ESIMD/lsc/lsc_usm_2d.cpp SYCL/ESIMD/lsc/lsc_slm_block_store.cpp SYCL/ESIMD/lsc/lsc_usm_prefetch_u8u32.cpp SYCL/ESIMD/lsc/lsc_surf_load_u8u32.cpp SYCL/ESIMD/lsc/lsc_surf_prefetch_u32.cpp SYCL/ESIMD/lsc/lsc_usm_prefetch_u32.cpp SYCL/ESIMD/lsc/lsc_surf_store_u16u32.cpp SYCL/ESIMD/lsc/lsc_surf_prefetch_u8u32.cpp SYCL/ESIMD/lsc/lsc_usm_store_u16u32.cpp SYCL/ESIMD/lsc/lsc_usm_load_u32.cpp SYCL/ESIMD/lsc/lsc_slm_block_load.cpp SYCL/ESIMD/lsc/lsc_usm_load_u16u32.cpp SYCL/ESIMD/lsc/lsc_surf_load_stateless.cpp SYCL/ESIMD/lsc/lsc_usm_load_u8u32.cpp SYCL/ESIMD/lsc/lsc_block_load_u32.cpp SYCL/ESIMD/lsc/lsc_usm.cpp SYCL/ESIMD/lsc/lsc_surf_load_u8_u16.cpp SYCL/ESIMD/lsc/lsc_slm.cpp SYCL/ESIMD/lsc/lsc_block_load_u16.cpp SYCL/ESIMD/lsc/lsc_block_prefetch_u16.cpp SYCL/ESIMD/lsc/lsc_slm_gather.cpp SYCL/ESIMD/lsc/lsc_usm_load_u64.cpp SYCL/ESIMD/lsc/lsc_usm_store_u64.cpp SYCL/ESIMD/lsc/lsc_block_store_u8.cpp SYCL/ESIMD/lsc/lsc_surf_store_u64.cpp SYCL/ESIMD/lsc/lsc_block_store_u16.cpp SYCL/ESIMD/lsc/lsc_usm_prefetch_u16u32.cpp SYCL/ESIMD/lsc/lsc_block_prefetch_u32.cpp SYCL/ESIMD/lsc/lsc_surf_load_u32.cpp SYCL/ESIMD/lsc/lsc_surf_store_u32.cpp SYCL/ESIMD/lsc/lsc_block_prefetch_u8.cpp SYCL/ESIMD/lsc/lsc_fence.cpp SYCL/ESIMD/lsc/lsc_usm_load_u8_u16.cpp SYCL/ESIMD/lsc/lsc_surf_store_u8u32.cpp SYCL/ESIMD/lsc/lsc_surf_store_stateless.cpp SYCL/ESIMD/lsc/lsc_usm_store_u32.cpp SYCL/ESIMD/histogram.cpp SYCL/ESIMD/private_memory/pm_access_3.cpp SYCL/ESIMD/private_memory/pm_access_2.cpp SYCL/ESIMD/private_memory/pm_access_1.cpp SYCL/ESIMD/vadd_2d_acc.cpp SYCL/ESIMD/slm_split_barrier.cpp SYCL/ESIMD/noinline_call_recursive.cpp SYCL/ESIMD/acc_gather_scatter_rgba_stateless.cpp SYCL/ESIMD/histogram_2d.cpp SYCL/ESIMD/PrefixSum.cpp SYCL/ESIMD/kmeans/kmeans.cpp SYCL/ESIMD/dword_atomic_smoke.cpp SYCL/ESIMD/accessor_load_store.cpp SYCL/ESIMD/vadd_half.cpp SYCL/ESIMD/linear/linear.cpp SYCL/ESIMD/matrix_transpose_usm.cpp SYCL/ESIMD/test_id_3d.cpp SYCL/ESIMD/BitonicSortKv2.cpp SYCL/ESIMD/vadd_usm_opqptr.cpp SYCL/ESIMD/Stencil.cpp SYCL/ESIMD/sycl_esimd_mix.cpp SYCL/ESIMD/histogram_256_slm.cpp SYCL/HostInteropTask/interop-task.cpp SYCL/Basic/code_location_e2e.cpp SYCL/Basic/image/image_accessor_readwrite_half.cpp SYCL/Basic/image/image_max_size.cpp SYCL/Basic/image/image_write_fp16.cpp SYCL/Basic/image/image_accessor_range.cpp SYCL/Basic/image/image_write.cpp SYCL/Basic/image/srgba-read.cpp SYCL/Basic/image/image_accessor_readwrite.cpp SYCL/Basic/image/image_accessor_readsampler.cpp SYCL/Basic/query_emulate_subdevice.cpp SYCL/Basic/event_profiling_info.cpp SYCL/Basic/kernel_bundle/kernel_bundle_api.cpp SYCL/Basic/intel-ext-device.cpp SYCL/Basic/handler/handler_mem_op.cpp SYCL/Basic/free_function_queries/free_function_queries_sub_group.cpp SYCL/Basic/built-ins.cpp SYCL/SubGroup/scan_spirv13_fp16.cpp SYCL/SubGroup/reduce_spirv13_fp16.cpp SYCL/SubGroup/scan_spirv13_fp64.cpp SYCL/SubGroup/reduce_spirv13_fp64.cpp SYCL/Config/kernel_from_file.cpp SYCL/DeviceCodeSplit/aot-gpu.cpp SYCL/Regression/implicit_kernel_bundle_image_filtering.cpp SYCL/KernelAndProgram/cache-build-result.cpp SYCL/KernelAndProgram/undefined-symbol.cpp SYCL/KernelAndProgram/build-log.cpp SYCL/InvokeSimd/invoke_simd_conv.cpp SYCL/InvokeSimd/invoke_simd_smoke.cpp

Here is an updated list of the tests marked as XFAIL and UNSUPPORTED for CUDA, which have been enabled, ran, and resulted in failures:

[+] XFAIL

N/A (All tests marked as XFAIL, which have been enabled, ran, and failed)

[+] UNSUPPORTED AOT/gpu.cpp BFloat16/bfloat16_type.cpp Basic/code_location_e2e.cpp Basic/image/image_accessor_readsampler.cpp Basic/image/image_accessor_readwrite.cpp Basic/image/image_accessor_readwrite_half.cpp Basic/image/image_max_size.cpp Basic/image/srgba-read.cpp Config/kernel_from_file.cpp DeviceCodeSplit/aot-gpu.cpp DeviceLib/imf_fp32_test.cpp DeviceLib/imf_fp64_test.cpp DeviceLib/imf_simd_emulate_test.cpp DiscardEvents/discard_events_using_assert.cpp ESIMD/BitonicSortK.cpp ESIMD/BitonicSortKv2.cpp ESIMD/Prefix_Local_sum1.cpp ESIMD/Prefix_Local_sum2.cpp ESIMD/Prefix_Local_sum3.cpp ESIMD/Stencil.cpp ESIMD/acc_gather_scatter_rgba.cpp ESIMD/acc_gather_scatter_rgba_stateless.cpp ESIMD/accessor.cpp ESIMD/accessor_gather_scatter.cpp ESIMD/accessor_gather_scatter_stateless.cpp ESIMD/accessor_load_store.cpp ESIMD/accessor_load_store_stateless.cpp ESIMD/accessor_stateless.cpp ESIMD/aot_mixed.cpp ESIMD/api/ballot.cpp ESIMD/api/bin_and_cmp_ops_heavy.cpp ESIMD/api/esimd_bit_ops.cpp ESIMD/api/esimd_merge.cpp ESIMD/api/esimd_pack_unpack_mask.cpp ESIMD/api/replicate_smoke.cpp ESIMD/api/saturation_smoke.cpp ESIMD/api/simd_any_all.cpp ESIMD/api/simd_binop_integer_promotion.cpp ESIMD/api/simd_copy_to_from.cpp ESIMD/api/simd_mask.cpp ESIMD/api/simd_memory_access.cpp ESIMD/api/simd_negation_operator.cpp ESIMD/api/simd_view_copy_move_assign.cpp ESIMD/api/simd_view_negation_operator.cpp ESIMD/api/simd_view_select_2d_fp.cpp ESIMD/api/simd_view_select_2d_int.cpp ESIMD/api/slm_gather_scatter_heavy.cpp ESIMD/api/svm_gather_scatter.cpp ESIMD/api/unary_ops_heavy.cpp ESIMD/dword_atomic_cmpxchg.cpp ESIMD/dword_atomic_smoke.cpp ESIMD/esimd_check_vc_codegen.cpp ESIMD/ext_math.cpp ESIMD/fp_args_size/fp_args_char_int_size_192.cpp ESIMD/fp_args_size/fp_args_char_int_size_256.cpp ESIMD/fp_args_size/fp_args_char_int_size_512.cpp ESIMD/fp_args_size/fp_args_char_int_size_64.cpp ESIMD/fp_args_size/fp_args_char_int_size_96.cpp ESIMD/fp_args_size/fp_args_int_size_192.cpp ESIMD/fp_args_size/fp_args_int_size_256.cpp ESIMD/fp_args_size/fp_args_int_size_512.cpp ESIMD/fp_args_size/fp_args_int_size_64.cpp ESIMD/fp_args_size/fp_args_int_size_96.cpp ESIMD/fp_in_phi.cpp ESIMD/histogram.cpp ESIMD/histogram_256_slm.cpp ESIMD/histogram_256_slm_spec_2020.cpp ESIMD/histogram_2d.cpp ESIMD/kmeans/kmeans.cpp ESIMD/linear/linear.cpp ESIMD/mandelbrot/mandelbrot.cpp ESIMD/mandelbrot/mandelbrot_spec.cpp ESIMD/matrix_transpose_glb.cpp ESIMD/matrix_transpose_usm.cpp ESIMD/noinline_args_size/noinline_args_char_int_size_192.cpp ESIMD/noinline_args_size/noinline_args_char_int_size_256.cpp ESIMD/noinline_args_size/noinline_args_char_int_size_512.cpp ESIMD/noinline_args_size/noinline_args_char_int_size_64.cpp ESIMD/noinline_args_size/noinline_args_char_int_size_96.cpp ESIMD/noinline_args_size/noinline_args_int_size_192.cpp ESIMD/noinline_args_size/noinline_args_int_size_256.cpp ESIMD/noinline_args_size/noinline_args_int_size_512.cpp ESIMD/noinline_args_size/noinline_args_int_size_64.cpp ESIMD/noinline_args_size/noinline_args_int_size_96.cpp ESIMD/preemption.cpp ESIMD/printf.cpp ESIMD/radix_sort.cpp ESIMD/reduction.cpp ESIMD/regression/abs_fix_test.cpp ESIMD/regression/big_const_initializer.cpp ESIMD/regression/bit_shift_compilation_test.cpp ESIMD/regression/complex-lib-lin.cpp ESIMD/regression/copyto_char_test.cpp ESIMD/regression/dgetrf.cpp ESIMD/regression/dgetrf_8x8.cpp ESIMD/regression/dgetrf_ref.cpp ESIMD/regression/fmod_compatibility_test.cpp ESIMD/regression/globals.cpp ESIMD/regression/half_conversion_test.cpp ESIMD/regression/operator_decrement.cpp ESIMD/regression/replicate_bug.cpp ESIMD/regression/store_zero_const.cpp ESIMD/regression/sycl_esimd_mixed_unnamed.cpp ESIMD/regression/unused_load.cpp ESIMD/regression/usm_gather_scatter_32.cpp ESIMD/regression/variable_gather_mask.cpp ESIMD/slm_barrier.cpp ESIMD/slm_split_barrier.cpp ESIMD/spec_const/spec_const_bool.cpp ESIMD/spec_const/spec_const_char.cpp ESIMD/spec_const/spec_const_double.cpp ESIMD/spec_const/spec_const_float.cpp ESIMD/spec_const/spec_const_int.cpp ESIMD/spec_const/spec_const_int64.cpp ESIMD/spec_const/spec_const_short.cpp ESIMD/spec_const/spec_const_uchar.cpp ESIMD/spec_const/spec_const_uint.cpp ESIMD/spec_const/spec_const_uint64.cpp ESIMD/spec_const/spec_const_ushort.cpp ESIMD/spec_const/unused_spec_const.cpp ESIMD/stencil2.cpp ESIMD/sycl_esimd_mix.cpp ESIMD/test_id_3d.cpp ESIMD/usm_gather_scatter_rgba.cpp ESIMD/vadd_1d.cpp ESIMD/vadd_2d.cpp ESIMD/vadd_2d_acc.cpp ESIMD/vadd_half.cpp ESIMD/vadd_usm.cpp ESIMD/vadd_usm_opqptr.cpp InvokeSimd/invoke_simd_conv.cpp InvokeSimd/invoke_simd_smoke.cpp KernelAndProgram/build-log.cpp KernelAndProgram/cache-build-result.cpp KernelAndProgram/undefined-symbol.cpp Matrix/element_wise_all_ops_cuda.cpp Matrix/joint_matrix_tensorcore.cpp Plugin/sycl-ls-gpu-default-any.cpp Regression/fsycl-host-compiler.cpp SeparateCompile/test.cpp SubGroup/reduce_spirv13_fp16.cpp SubGroup/scan_spirv13_fp16.cpp

Finally, the following list is the tests marked as UNSUPPORTED, which have been enabled, ran, and resulted in success:

SYCL/Assert/assert_in_simultaneous_kernels.cpp SYCL/Assert/assert_in_simultaneously_multiple_tus.cpp SYCL/Assert/assert_in_simultaneously_multiple_tus_one_ndebug.cpp SYCL/Basic/built-ins.cpp SYCL/Basic/event_profiling_info.cpp SYCL/Basic/handler/handler_mem_op.cpp SYCL/Basic/image/image_write.cpp SYCL/Basic/image/image_write_fp16.cpp SYCL/Basic/intel-ext-device.cpp SYCL/Basic/kernel_bundle/kernel_bundle_api.cpp SYCL/Basic/query_emulate_subdevice.cpp SYCL/DeprecatedFeatures/deprecated_intel_ext_device.cpp SYCL/DeviceLib/ITTAnnotations/atomic.cpp SYCL/DeviceLib/ITTAnnotations/barrier.cpp SYCL/ESIMD/api/esimd_rgba_smoke.cpp SYCL/ESIMD/api/simd_subscript_operator.cpp SYCL/ESIMD/api/simd_view_subscript_operator.cpp SYCL/ESIMD/api/slm_gather_scatter.cpp SYCL/ESIMD/api/slm_gather_scatter_rgba.cpp SYCL/ESIMD/PrefixSum.cpp SYCL/ESIMD/vec_arg_call_conv_ext.cpp SYCL/ESIMD/vec_arg_call_conv_smoke.cpp SYCL/InlineAsm/asm_16_empty.cpp SYCL/InlineAsm/asm_16_matrix_mult.cpp SYCL/InlineAsm/asm_16_no_input_int.cpp SYCL/InlineAsm/asm_16_no_opts.cpp SYCL/InlineAsm/asm_8_empty.cpp SYCL/InlineAsm/asm_8_no_input_int.cpp SYCL/InlineAsm/asm_arbitrary_ops_order.cpp SYCL/InlineAsm/asm_decl_in_scope.cpp SYCL/InlineAsm/asm_float_add.cpp SYCL/InlineAsm/asm_float_imm_arg.cpp SYCL/InlineAsm/asm_float_neg.cpp SYCL/InlineAsm/asm_if.cpp SYCL/InlineAsm/asm_imm_arg.cpp SYCL/InlineAsm/asm_loop.cpp SYCL/InlineAsm/asm_mul.cpp SYCL/InlineAsm/asm_multiple_instructions.cpp (SHOULD HAVE FAILED, REQUIRES GPU; XFAIL GPU) SYCL/InlineAsm/asm_no_operands.cpp SYCL/InlineAsm/asm_no_output.cpp SYCL/InlineAsm/asm_plus_mod.cpp SYCL/InlineAsm/asm_switch.cpp SYCL/InlineAsm/letter_example.cpp SYCL/InlineAsm/malloc_shared_32.cpp SYCL/InlineAsm/malloc_shared_in_out_dif.cpp SYCL/InlineAsm/malloc_shared_no_input.cpp SYCL/InlineAsm/Negative/asm_bad_opcode.cpp SYCL/InlineAsm/Negative/asm_bad_operand_syntax.cpp SYCL/InlineAsm/Negative/asm_duplicate_label.cpp SYCL/InlineAsm/Negative/asm_illegal_exec_size.cpp SYCL/InlineAsm/Negative/asm_missing_label.cpp SYCL/InlineAsm/Negative/asm_missing_region.cpp SYCL/InlineAsm/Negative/asm_simple.cpp SYCL/InlineAsm/Negative/asm_undefined_decl.cpp SYCL/InlineAsm/Negative/asm_undefined_pred.cpp SYCL/InlineAsm/Negative/asm_wrong_declare.cpp SYCL/Plugin/sycl-ls.cpp SYCL/Regression/implicit_kernel_bundle_image_filtering.cpp SYCL/SubGroup/reduce_spirv13_fp64.cpp SYCL/SubGroup/scan_spirv13_fp64.cpp SYCL/USM/dep_events.cpp SYCL/USM/prefetch.cpp

AerialMantis commented 2 years ago

We've re-evaluated all the current XFAIL and UNSUPPORTED tests for the CUDA backend, tests which were marked as UNSUPPORTED but were actually passing have been re-enabled in https://github.com/intel/llvm-test-suite/pull/1317 and some tests which were marked as XFAIL which should have been UNSUPPORTED have been updated in https://github.com/intel/llvm-test-suite/pull/1340.

We are investigate all remaining XFAIL and UNSUPPORTED, so we will close this ticket now.

intel / llvm-test-suite

XFAIL: cuda tracker #249