Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/apache/tvm-ffi
/ functions
Functions
5,333 in github.com/apache/tvm-ffi
⨍
Functions
5,333
◇
Types & classes
1,261
↳
Endpoints
2
↓ 1 callers
Function
raise_cxx_error
()
tests/python/test_error.py:106
↓ 1 callers
Function
raise_error
()
tests/python/test_error.py:65
↓ 1 callers
Function
register
Register the error class name with the FFI core.
python/tvm_ffi/error.py:247
↓ 1 callers
Function
register
Register the global function with the FFI core.
python/tvm_ffi/registry.py:167
↓ 1 callers
Function
register_object
Register object type. Parameters ---------- type_key The type key of the node. It requires ``type_key`` to be registered already
python/tvm_ffi/registry.py:36
↓ 1 callers
Function
remove_global_func
Remove a global function by name. Parameters ---------- name The name of the global function. Examples -------- .. c
python/tvm_ffi/registry.py:237
↓ 1 callers
Method
reserve
! * \brief Make sure the list has the capacity of at least n * \param n lower bound of the capacity */
include/tvm/ffi/container/array.h:532
↓ 1 callers
Method
reset
! \brief reset the content of ptr to be nullptr */
include/tvm/ffi/object.h:568
↓ 1 callers
Method
reset
! \brief reset the content of ptr to be nullptr */
include/tvm/ffi/object.h:715
↓ 1 callers
Function
run_add_one
Invoke add_one from the extension and print the result.
examples/python_packaging/run_example.py:30
↓ 1 callers
Function
run_add_two
Invoke add_two from the extension and print the result.
examples/python_packaging/run_example.py:24
↓ 1 callers
Function
run_benchmark
Run overhead benchmarks for both Triton and TVM-FFI. Parameters ---------- cubin_bytes : bytes Compiled CUBIN bytes num_calls
examples/cubin_launcher/benchmark_overhead.py:212
↓ 1 callers
Function
run_int_pair
Invoke IntPair from the extension to demonstrate object handling.
examples/python_packaging/run_example.py:45
↓ 1 callers
Function
run_parallel
(cmd: list[str], files: list[str], jobs: int)
tests/lint/clang_tidy_precommit.py:119
↓ 1 callers
Function
run_raise_error
Invoke raise_error from the extension to demonstrate error handling.
examples/python_packaging/run_example.py:36
↓ 1 callers
Method
same_as
! * \brief Comparator * \param other Another object ref. * \return the compare result. */
include/tvm/ffi/object.h:783
↓ 1 callers
Method
shape
! * \brief Get the shape of the Tensor. * \return The shape of the Tensor. */
include/tvm/ffi/container/tensor.h:311
↓ 1 callers
Method
shape
! * \brief Get the shape of the Tensor. * \return The shape of the Tensor. */
include/tvm/ffi/container/tensor.h:750
↓ 1 callers
Function
should_skip_file
Check if file should be skipped based on SKIP_LIST.
tests/lint/check_asf_header.py:177
↓ 1 callers
Method
sum
(self, /)
python/tvm_ffi/testing/testing.py:94
↓ 1 callers
Method
sum
(self, /)
examples/python_packaging/python/my_ffi_extension/_ffi_api.py:55
↓ 1 callers
Function
test_echo
()
tests/python/test_function.py:37
↓ 1 callers
Method
test_map
tvm_ffi.Map → canonical origin 'Map'.
tests/python/test_type_converter.py:4173
↓ 1 callers
Function
test_raise_error
(_0: str, _1: str, /)
python/tvm_ffi/testing/_ffi_api.py:94
↓ 1 callers
Function
toposort_objects
Collect ObjectInfo objects for type keys, topologically sorted by inheritance.
python/tvm_ffi/stub/lib_state.py:74
↓ 1 callers
Function
trigger_cpp_side_error
()
tests/python/test_error.py:134
↓ 1 callers
Function
tvm_ffi_nop
Overhead of tvm FFI python call via calling a NOP. testing.nop is defined in c++ and do nothing.
tests/scripts/benchmark_dlpack.py:130
↓ 1 callers
Function
tvm_ffi_nop_autodlpack_from_numpy
Measures overhead of running dlpack via auto convert by directly take numpy.ndarray as inputs.
tests/scripts/benchmark_dlpack.py:275
↓ 1 callers
Function
tvm_ffi_nop_from_numpy_dlpack
Run dlpack conversion + tvm_ffi.nop. Measures overhead of running dlpack for each args then invoke
tests/scripts/benchmark_dlpack.py:179
↓ 1 callers
Function
tvm_ffi_nop_from_torch_dlpack
Run dlpack conversion + tvm_ffi.nop. Measures overhead of running dlpack for each args then invoke
tests/scripts/benchmark_dlpack.py:168
↓ 1 callers
Function
tvm_ffi_nop_from_torch_utils_to_dlpack
Measures overhead of running dlpack for each args then invoke but uses the legacy torch.utils.dlpack.to_dlpack API. This helps to measure pos
tests/scripts/benchmark_dlpack.py:201
↓ 1 callers
Function
tvm_ffi_self_dlpack_nop
Run dlpack conversion + tvm_ffi.nop. Measures overhead of running dlpack for each args then invoke
tests/scripts/benchmark_dlpack.py:190
↓ 1 callers
Method
type_index
! \return The internal runtime type index of the object. */
include/tvm/ffi/object.h:228
↓ 1 callers
Method
update
Update the dict from a mapping.
python/tvm_ffi/container.py:771
↓ 1 callers
Function
update_ld_library_path
Update the LD_LIBRARY_PATH environment variable so cargo run/test can directly pick it up note that it won't always work later consumption of the libr
rust/tvm-ffi-sys/build.rs:26
↓ 1 callers
Function
update_ld_library_path
Update the LD_LIBRARY_PATH environment variable so cargo run/test can directly pick it up note that it won't always work later consumption of the libr
rust/tvm-ffi/build.rs:41
↓ 1 callers
Function
use_cubin_kernel
Load and test CUBIN kernels through TVM-FFI. Parameters ---------- cubin_bytes : bytes Compiled CUBIN bytes Returns ----
examples/cubin_launcher/example_nvrtc_cubin.py:79
↓ 1 callers
Function
use_cubin_kernel
Load and test Triton CUBIN kernel through TVM-FFI. Parameters ---------- cubin_bytes : bytes Compiled CUBIN bytes Returns
examples/cubin_launcher/example_triton_cubin.py:79
↓ 1 callers
Method
value_or
include/tvm/ffi/optional.h:313
↓ 1 callers
Function
weak_count
(handle: *mut TVMFFIObject)
rust/tvm-ffi/src/object.rs:195
↓ 1 callers
Function
with_stream
Call `f` with the device stream temporarily set to `stream`. # Safety `stream` must be a valid stream handle for the given device, or null.
rust/tvm-ffi/src/device.rs:34
↓ 1 callers
Function
worker
Worker function that tries to acquire lock and increment counter.
tests/python/utils/filelock_worker.py:25
↓ 1 callers
Method
~Any
! * \brief Destructor */
include/tvm/ffi/any.h:267
Method
AccessPath
include/tvm/ffi/reflection/access_path.h:367
Method
AccessPathObj
default constructor to enable auto-serialization
include/tvm/ffi/reflection/access_path.h:205
Method
AccessStep
! * \brief Constructor * \param kind The kind of the access step. * \param key The key of the access step. */
include/tvm/ffi/reflection/access_path.h:113
Method
AccessStepObj
default constructor to enable auto-serialization
include/tvm/ffi/reflection/access_path.h:80
Method
Add
tests/cpp/testing_object.h:118
Method
AddI64
src/ffi/testing/testing.cc:114
Method
AddInt
Some methods to exercise RegisterMethod
src/ffi/testing/testing.cc:749
Method
AddObjectFile
addons/tvm_ffi_orcjit/src/ffi/orcjit_dylib.cc:107
Function
AddOne
! * \brief Launch add_one_cuda kernel on input tensor. * \param x Input tensor (float32, 1D) * \param y Output tensor (float32, 1D, same shape as x
examples/cubin_launcher/embedded_cubin/cpp_embed/src/lib_embedded.cc:49
Function
AddOne
! * \brief Launch add_one_cuda kernel on input tensor. * \param x Input tensor (float32, 1D) * \param y Output tensor (float32, 1D, same shape as x
examples/cubin_launcher/embedded_cubin/embed_with_tvm_ffi/src/lib_embedded.cc:46
Function
AddOne
! * \brief Launch add_one_cuda kernel on input tensor. * \param x Input tensor (float32, 1D) * \param y Output tensor (float32, 1D, same shape as x
examples/cubin_launcher/embedded_cubin/include_bin2c/src/lib_embedded.cc:46
Function
AddOne
! * \brief Launch add_one_cuda kernel on input tensor. * \param x Input tensor (float32, 1D) * \param y Output tensor (float32, 1D, same shape as x
examples/cubin_launcher/dynamic_cubin/src/lib_dynamic.cc:61
Function
AddOne
examples/python_packaging/src/extension.cc:38
Function
AddOne
! \brief Perform vector add one: y = x + 1 (1-D float32) */
examples/quickstart/compile/add_one_cpu.cc:27
Method
AddOneFloat
tests/cpp/test_overload.cc:38
Method
AddOneInt
tests/cpp/test_overload.cc:37
Function
AddTwo
[tvm_ffi_abi.begin]
examples/python_packaging/src/extension.cc:32
Method
Agregate
include/tvm/ffi/container/array.h:680
Method
AgregateImpl
include/tvm/ffi/container/array.h:562
Function
AlignedAlloc
include/tvm/ffi/memory.h:56
Method
AllocData
tests/cpp/test_tensor.cc:27
Method
AllocData
tests/cpp/test_example.cc:127
Method
AllocData
tests/cpp/extra/test_c_env_api.cc:36
Method
AllocData
examples/quickstart/load/load_cpp.cc:50
Method
AllocData
examples/quickstart/load/load_cuda.cc:50
Method
Any
! * \brief Default constructor */
include/tvm/ffi/any.h:259
Function
Any_AnyView_Destroy
[Any_AnyView.Destroy.begin]
examples/abi_overview/example_code.c:76
Function
Any_AnyView_FromFloat
examples/abi_overview/example_code.c:51
Function
Any_AnyView_FromObjectPtr
[Any_AnyView.FromObjectPtr.begin]
examples/abi_overview/example_code.c:61
Function
Any_AnyView_GetDLTensor
[Any_AnyView.GetDLTensor.begin]
examples/abi_overview/example_code.c:111
Function
Any_AnyView_GetFloat
examples/abi_overview/example_code.c:99
Function
Any_AnyView_GetObject
[Any_AnyView.GetObject.begin]
examples/abi_overview/example_code.c:123
Method
AppendInt
src/ffi/testing/testing.cc:750
Method
Apply
include/tvm/ffi/function_details.h:47
Method
Apply
! * \brief Move metadata into `FieldInfoBuilder` * \param info The field info builder. */
include/tvm/ffi/reflection/registry.h:90
Method
Apply
! * \brief Apply the default value to the field info * \param info The field info. */
include/tvm/ffi/reflection/registry.h:156
Method
Apply
! * \brief Apply the default factory to the field info * \param info The field info. */
include/tvm/ffi/reflection/registry.h:189
Method
Apply
! * \brief Apply the field flag to the field info * \param info The field info. */
include/tvm/ffi/reflection/registry.h:247
Method
Apply
! * \brief Apply the compare flag to the field info. * \param info The field info. */
include/tvm/ffi/reflection/registry.h:298
Method
Apply
! * \brief Apply the hash flag to the field info. * \param info The field info. */
include/tvm/ffi/reflection/registry.h:326
Method
Apply
! * \brief Apply the kw_only flag to the field info. * \param info The field info. */
include/tvm/ffi/reflection/registry.h:358
Method
Apply
! * \brief Apply the init flag to the field info. * \param info The field info. */
include/tvm/ffi/reflection/registry.h:673
Method
ApplyExtraInfoTrait
include/tvm/ffi/reflection/registry.h:443
Method
ApplyFieldInfoTrait
include/tvm/ffi/reflection/registry.h:423
Method
ApplyMethodInfoTrait
include/tvm/ffi/reflection/registry.h:433
Function
ArrayContains
(_0: Sequence[Any], _1: Any, /)
python/tvm_ffi/_ffi_api.py:40
Function
ArrayGetItem
(_0: Sequence[Any], _1: int, /)
python/tvm_ffi/_ffi_api.py:41
Method
ArrayItem
src/ffi/extra/visit_error_context.cc:112
Method
ArrayItem
! * \brief Create an access step for a array item access. * \param index The index of the array item to access. * \return The access step.
include/tvm/ffi/reflection/access_path.h:139
Method
ArrayItemMissing
! * \brief Create an access step for a array item missing access. * \param index The index of the array item to access. * \return The access
include/tvm/ffi/reflection/access_path.h:146
Function
ArraySize
(_0: Sequence[Any], /)
python/tvm_ffi/_ffi_api.py:42
Method
Assign
include/tvm/ffi/container/array.h:632
Method
Assign
include/tvm/ffi/container/list.h:449
Method
AttachFieldFlag
! * \brief Attach a field flag to the field * \param flag The flag to be set */
include/tvm/ffi/reflection/registry.h:213
Method
Attr
src/ffi/extra/visit_error_context.cc:94
Method
Attr
! * \brief Create an access step for a object attribute access. * \param field_name The name of the field to access. * \return The access ste
include/tvm/ffi/reflection/access_path.h:121
Method
AttrMissing
! * \brief Create an access step for a object attribute missing access. * \param field_name The name of the field to access. * \return The ac
include/tvm/ffi/reflection/access_path.h:130
← previous
next →
1,001–1,100 of 5,333, ranked by callers