### Description
Add new name "WebGPU_Buffer" to OrtMemoryInfo.
This is one of the prerequisites for supporting IO binding for WebGPU
buffer in onnxruntime-web.
list of prerequisites PRs:
#17465#17469 (this one)
### Description
Remove AllocatorManager class
### Motivation and Context
After the refactor PR #15833 is in, AllocatorManager class is not
referenced anymore.
### Description
Run clang-format in CI. Formatted all c/c++, objective-c/c++ files.
Excluded
```
'onnxruntime/core/mlas/**',
'onnxruntime/contrib_ops/cuda/bert/tensorrt_fused_multihead_attention/**',
```
because they contain assembly or is data heavy
### Motivation and Context
Coding style consistency
### Description
This PR is to address follow-up comments for the multi-stream pr
https://github.com/microsoft/onnxruntime/pull/13495
Changes including:
- Make StreamAwareArena transparent to minimal build
- Make DeviceStreamCollection transparent to minimal build
- Replace ORT_MUST_USE_RESULT with [[nodiscard]]
- Remove unnecessary shared_ptr
### Motivation and Context
This PR is to address follow-up comments for the multi-stream pr
https://github.com/microsoft/onnxruntime/pull/13495
Co-authored-by: Lei Cao <leca@microsoft.com>
**Description**: This PR including following works:
1. provide stream and related synchronization abstractions in
onnxruntime.
2. enhance onnxruntime's execution planner / executor / memory arena to
support execute multiple streams in parallel.
3. deprecate the parallel executor for cpu.
4. deprecate the Fence mechanism.
5. update the cuda / tensorrt EP to support the stream mechanism,
support running different request in different cuda stream.
**Motivation and Context**
- Why is this change required?
currently, the execution plan is just a linear list of those primitives,
ort will execute them step by step. For any given graph, ORT will
serialize it to a fixed execution order. This sequential execution
design simplifies most scenarios, but it has the following limitations:
1. it is difficult to enable inter-node parallelization, we have a
half-baked parallel executor but it is very difficult to make it work
with GPU.
2. The fence mechanism can work with single gpu stream + cpu thread
case, but when extend to multiple stream, it is difficult to manage the
cross GPU stream synchronizations.
3. our cuda EP rely on the BFCArena to make the memory management work
with the GPU async kernels, but current BFCArena is not aware of the
streams, so it doesn't behavior correctly when run with multiple
streams.
This PR enhance our existing execution plan and executor to support
multiple stream execution. we use an unified algorithm to mange both
single stream and multiple stream scenarios.
This PR mainly focus on the infrastructure support for multiple stream
execution, that is said, given a valid stream assignment, onnxruntime
can execute it correctly. How to generate a good stream assignment for a
given model will be in the future PR.
Co-authored-by: Cheng Tang <chenta@microsoft.com@orttrainingdev9.d32nl1ml4oruzj4qz3bqlggovf.px.internal.cloudapp.net>
Co-authored-by: Cheng Tang <chenta@microsoft.com>
Co-authored-by: RandySheriffH <48490400+RandySheriffH@users.noreply.github.com>
Co-authored-by: Randy Shuai <rashuai@microsoft.com>
Co-authored-by: cao lei <jslhcl@gmail.com>
Co-authored-by: Lei Cao <leca@microsoft.com>
### Description
Deprecate CustomOpApi and refactor dependencies for exception safety and
eliminate memory leaks.
Refactor API classes for clear ownership and semantics.
Introduce `InitProviderOrtApi()`
### Motivation and Context
Make public API better and safer.
Special note about `Ort::Unowned`. The class suffers from the following
problems:
1. It is not able to hold const pointers to the underlying C objects.
This forces users to `const_cast` and circumvent constness of the
returned object. The user is now able to call mutating interfaces on the
object which violates invariants and may be a thread-safety issue. It
also enables to take ownership of the pointer and destroy it
unintentionally (see examples below).
2. The objects that are unowned cannot be copied and that makes coding
inconvenient and at times unsafe.
3. It directly inherits from the type it `unowns`.
All of the above creates great conditions for inadvertent unowned object
mutations and destructions. Consider the following examples of object
slicing, one of them is from a real customer issue and the other one I
accidentally coded myself (and I am supposed to know how this works).
None of the below can be solved by aftermarket patches and can be hard
to diagnose.
#### Example 1 slicing of argument
```cpp
void SlicingOnArgument(Ort::Value& value) {
// This will take possession of the input and if the argument
// is Ort::Unowned<Ort::Value> it would again double free the ptr
// regardless if it was const or not since we cast it away.
Ort::Value output_values[] = {std::move(value)};
}
void main() {
const OrtValue* ptr = nullptr; // some value does not matter
Ort::Unowned<Ort::Value> unowned{const_cast<OrtValue*>(ptr)};
// onowned is destroyed when the call returns.
SlicingOnArgument(unowned);
}
```
#### Example 2 slicing of return value
```cpp
// The return will be sliced to Ort::Value that would own and relase (double free the ptr)
Ort::Value SlicingOnReturn() {
const OrtValue* ptr = nullptr; // some value does not matter
Ort::Unowned<Ort::Value> unowned{const_cast<OrtValue*>(ptr)};
return unowned;
}
```
Adding code for symmetric quantized matrix multiplication. Used in quantized convolution, achieving significant perf gain.
TODO, use Symmetric Quantized GEMM in other operators!
TODO address activation buffer overread in custom allocators and tensors supplied by users.
DOT kernel perf test:
Pixel 5a:
Cartoongan 513.539 ms 471.786 ms
Efficient 57.5169 ms 56.4174 ms
Edgetpu 14.6673 ms 13.5959 ms
NEON kernel perf test
Pixel 3a
Cartoongan 1423.53 ms 1069.92 ms
Efficient 114.086 ms 107.968 ms
Edgetpu 39.2632 ms 36.9839 ms
Co-authored-by: Chen Fu <fuchen@microsoft.com>
* Added checks for Hetero/Multi
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Remote Context Plugin
* changes for IO Buffer plugin
* erronous couts added
* erronous entry rectified
* Set the Openvino OP Buffer also as output
* Enable AUTO plugin in OpenVINO EP
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Remote Context Plugin
* changes for IO Buffer plugin
* erronous couts added
* erronous entry rectified
* Added checks for Hetero/Multi
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Set the Openvino OP Buffer also as output
* Enable AUTO plugin in OpenVINO EP
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Please commit error message and rectification of param.context
* Alignment fixed
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Changed the string to OpenVINO_GPU
* hanged OpenVINO to to OpenVINO_CPU
* Onnxruntime updated API for memory location
* Removing Duplicate LOG Error
* Tensor.h removed DeviceType function. Updated comment
* API Comments updated
* Removing changes to Provider Indo
* Erronous commit
* Removing Extra logs
* Merge CMAKE
* Not copy from a local location
* Duplicate Entry
* Remove extra line
Co-authored-by: MaajidKhan <n.maajidkhan@gmail.com>
* Share allocator between CUDA EP & TRT EP.
limitation:
1. Does not cover the per-thread allocator created by CUDA EP, still need to figure out the way to remove it
2. Need to have more identifiers to make it able to share CPU allocator across all EPs
* Fix issue: https://github.com/microsoft/onnxruntime/issues/6094
Root cause: we didn't expose the OrtMemoryInfo for TRT, so it will cause issue if user want use IObinding for Tensorrt.
Short term fix, add the OrtMemoryInfo for TRT. Long term should unify the allocator for CUDA and TRT
- Update IAllocator setup to move the OrtMemoryInfo to the base class instead of requiring derived classes to have that as a member and override a virtual method to return it.
- Cleanup CreateAllocator setup to take an argument as to whether to wrap the device allocator in an arena allocator. The choice to do that isn't a property of the underlying device allocator.
- Minor cleanups in the various EPs to adjust to the change to IAllocator and CreateAllocator, and to use the create_arena flag consistently when available.
1. Fix static analysis warnings found by VC++
2. Add a new pipeline for static analysis
3. Merge all the windows CI build into one single yaml file.(Easier to queue them all).
4. Make DNNL build faster by disabling building the tests and examples.
5. Enable custom op unitest.
* port the mimalloc allocator
* hook mimalloc opt into common.h and reduction ops
* repurpose USE_MIMALLOC to only denote subbing in of default allocator with mimalloc and some refactoring
* fix unintended cherry pick diffs
* polish alloctor_mimalloc
* explicitly disable mimalloc where it already had been disabled
* update mimalloc to pull in stl allocator
* switch mimalloc stl allocator to use mimalloc library version
* turn mimalloc on by default (only the stl changes are enabled, the python interacting ones are off already and shall remain so)
* move FastAllocVector into cpu specific code
* separate out defines into arena and stl changes
* the rest of the define renames
* bfc arena allocator
* some typos and rename the bfc arena allocator to fit existing class naming conventions
* adjustments in response to comments
* different template instantiations are friends
* Remove allocator type from the key comparison in ExecutionProviders.
Remove usage of DummyArena as it's no longer necessary.
* Fix x86 tests where arena allocator is disabled.
Make initialization of OrtMemoryInfo clearer by adding Invalid enum value.
* Make OrtValueNameIdxMap::MaxIdx more intuitive.
Description: make default CPU allocator to use MLAS preferred alignment
Motivation and Context
This is needed for C API to have an aligned default CPU allocator, the same as the one in CPU provider
Add MlasGetPreferredBufferAlignment() for use by CPUAllocator::Alloc to get the byte alignment for CPU tensors. Using MLAS allows the value to be based on the platform the binary is running on instead of a constant value fixed at compile time.
* added packaging pipeline
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* put the c-api header file at root instead of under core/session
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* Update win-ci-pipeline.yml for Azure Pipelines
* parameterize the windows build script
* Update win-package-pipeline.yml for Azure Pipelines
* fixed indenting
* fixed indenting
* fix parameter reference syntax
* try using arch = amd64 for the vcvarsall
* remove duplicate tasks
* use vcvarsall
* some more refactor
* fix typo
* fix typo
* factored out the packaging step into a template
* add x86 build to package pipeline
* use amd64 for vcvars arg
* added gpu pipeline. added msbuild platform param
* fix the msbuild platform
* use amd64 host for x86 build
* use buildarch=x86 for vcvarsall
* remove vcvars from setup steps
* add some logging for PNG lib, and disable fns_candy demo for win32
* set allocator alignment to 32 bit for win32 compiler
* disable parallel execution test for x86
* use 64 bit toolchain for x86 build
* add missing -T flag for toolset
* fix string delimietr in workingdirectory name for package build test step
* fix gpu pipeline
* make io_types test conditional
* use cuda 10 instead of cuda 9.1, similar to the ci build
* try some workaround on the io test
* undo inadvertent local change in build.py, also reenable the io test
* make all test run single threaded
* blacklist few failing tests for x86
* added some log in build.py
* edit build.py to disable parallel test
* add the failed tests into the blacklist for win32
* add tf_pasnet_large to blacklist
* change control flow for build.py onnx tests
* add README, license and TPN to the package
* updated build.py test sequence for parallel executor
* updated onnx test flow as per review comment
* add type checking log in the compare_mlvalue
* fix type cast
* blacklist some failed test as of now
* one more blacklisted test
* merge function compile interface
* fix build error
* fix linux build break
* fix static cast issue; fix clang style
* fix argument change
* use alignment allocation;fix comments in pr
* fix linux break
* apply clang format
* rename according to comments in pr
* rename according to pr comments;remove useless file
* remove the need_compile flag
* avoid passing whole session state
Applies to all public headers and macros, plus many internal ones. There are still some internal things with OnnxRuntime in the name, but this fixes all public functions & macros.
