This PR applies clang-tidy readability checks to jit sources and all headers in the code base.
`readability-redundant-inline-specifier` is suppressed because it incurs too many changes. `readability-redundant-inline-specifier` is used to detect redundant inline specifiers on function and variable declarations. There are many in-class method definitions that are marked inline.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/164652
Approved by: https://github.com/Skylion007
This PR applies clang-tidy readability checks to jit sources and all headers in the code base.
`readability-redundant-inline-specifier` is suppressed because it incurs too many changes. `readability-redundant-inline-specifier` is used to detect redundant inline specifiers on function and variable declarations. There are many in-class method definitions that are marked inline.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/164652
Approved by: https://github.com/Skylion007
## Overview
This PR allows the profiler users to access `Kineto` and `TorchOp` metadata in JSON string format through a new `metadata_json` attribute in `FunctionEvent` objects, which is triggered through a new `expose_kineto_event_metadata` flag in `ExperimentalConfig`.
## Testing
A unit test was added to validate functionality.
## Documentation
Added/updated function doc strings where appropriate.
## Example output
```python
import torch
from torch.profiler import profile
with profile(experimental_config=torch._C._profiler._ExperimentalConfig(expose_kineto_event_metadata=True)) as prof:
res = torch.mm(torch.rand(1024, 1024), torch.rand(1024, 1024))
for event in prof.events():
print(f'name: {event.key}, metadata: {event.metadata_json}')
```
```
name: aten::rand, metadata: "Ev Idx": 0
name: aten::empty, metadata: "Ev Idx": 1
name: aten::uniform_, metadata: "Ev Idx": 2
name: aten::rand, metadata: "Ev Idx": 3
name: aten::empty, metadata: "Ev Idx": 4
name: aten::uniform_, metadata: "Ev Idx": 5
name: aten::mm, metadata: "Ev Idx": 6
name: aten::resolve_conj, metadata: "Ev Idx": 7
name: aten::resolve_conj, metadata: "Ev Idx": 8
name: aten::resolve_conj, metadata: "Ev Idx": 9
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/161624
Approved by: https://github.com/sraikund16
Summary:
Adds Python Garbage Collection to Kineto Traces and Profiler FunctionEvents. Create custom cpp callback in profiler_python.cpp. Then define a python function with cpp and register that callback for all python garbage collection. We don't worry about thread safety in this case because we are only doing init/teardown for main thread while holding GIL.
Currently we are hiding this behind experimental config because python tracing tends to be unstable especially when adding any new feature. If this is found to not add too much overhead we can set this to on by default. NOTE: To enable this you need both with_stack=True and the experimental config on!
Test Plan:
Ran trace with GC induced and saw it on trace
Also added a test
Rollback Plan:
Differential Revision: D80491146
Pull Request resolved: https://github.com/pytorch/pytorch/pull/161209
Approved by: https://github.com/ngimel
Currently, recorded profiler events for aten ops do not store overload names. It would be useful to know which overloads are actually called to analyse performance.
For example, consider the following dispatch trace which occurs if there is a fallthrough kernel registered for aten::add:
```
[call] op=[aten::add.Tensor], key=[AutogradCPU]
[redispatch] op=[aten::add.Tensor], key=[Undefined]
[call] op=[aten::empty.memory_format], key=[BackendSelect]
[redispatch] op=[aten::empty.memory_format], key=[CPU]
[call] op=[aten::add.out], key=[CPU]
```
In this case, aten::add.out is a child of aten::add.Tensor, however the current profiler trace provides no way to differentiate aten op calls.
See the added unit test for a more detailed example.
Fixes #ISSUE_NUMBER
Pull Request resolved: https://github.com/pytorch/pytorch/pull/143114
Approved by: https://github.com/sraikund16
Summary:
D62008788 added an extra parameter to the RawTensorMetadata struct. For some reason this causes some corrupted accesses in other tests as described in T200685032.
Once this is removed the tests pass. Going forward we need to document how to add parameters to this portion of the code as the AppendOnlyLists seem to be very rigid.
Test Plan: Ran all the tests locally and they all passed.
Differential Revision: D62171089
Pull Request resolved: https://github.com/pytorch/pytorch/pull/135096
Approved by: https://github.com/aaronenyeshi
Summary:
Currently some jobs are encountering the following trace, P1539415198. This suggests that when we are parsing through tensors the path is prone to encountering an invalid address. This is is possibly occurring because for some reason the sizes() and strides() of a Tensor seem to not be of the same dimensions. We assume such when iterating through the shapes to get the Ivalue generator. When browsing some of the tensor implementations, I found that some of the size and stride paths are different which could be the cause of this issue. Regardless, the profiler should be flexible enough to handle such issues without bringing down the whole main thread.
If the crashes still persist, it will still give us a data point as to where they are occurring and we can rule out the strides/sizes as the culprit
Test Plan: This change doesn't break anything in the happy path, just makes sure the bad path is not exited abruptly. We should use this in order to debug what the events are having mismatching dimensions between sizes and strides.
Differential Revision: D62008788
Pull Request resolved: https://github.com/pytorch/pytorch/pull/134862
Approved by: https://github.com/aaronenyeshi
Summary: On the autograd side of things, we are currently saving the kwinputs but we aren't doing anything with them on the profiler side. This diff enables the use of the kwinputs for both FunctionEvents and Chrome Traces.
Test Plan: Added unit testing for both chrome traces and FunctionEvents. Used RecordFunctionFast to test kwinputs since test already had kwargs being passed in but not tested.
Differential Revision: D59472345
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130373
Approved by: https://github.com/davidberard98
Summary:
Currently we have an issue where CPU User annotations can overlap with python events in the event that a python event calls step() within the function itself. To combat this, we can move the left side of the user annotation to the beginning of the parent python function. We do this because when instantiating the profiler we already start on step 0.
To implement this, we start by collecting all instances of ProfilerStep during post processing. Since TorchOps and Python events are sorted already, we can easily check if the current python event partially overlaps with the current ProfilerStep and, if so, alter the start time of the current ProfilerStep. We then move to the next ProfilerStep and continue iterating through all the python events. This keeps the time complexity of adding events to 'out' at O(s + n) -> O(n) post sorting, where "s" is the number of ProfilerSteps and "n" is the length of all events.
Test Plan:
Added unit test in which step() is called midway through a function. Afterwards, we print out a trace and then load the json to check that there are no overlaps. Also make sure that there is no regression in performance.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/129599
Approved by: https://github.com/aaronenyeshi
Summary:
Kineto traces use microsecond level granularity because of chrome tracing defaults to that precision. Fix by adding preprocessor flag to TARGETS and BUCK files. Also remove any unnecessary ns to us conversions made in the profiler itself.
This diff contains profiler changes only. Libkineto changes found in D54964435.
Test Plan:
Check JSON and chrome tracing to make sure values are as expected. Tracing with flags enabled should have ns precision. Tracings without flags should be same as master.
Zoomer: https://www.internalfb.com/intern/zoomer/?profiling_run_fbid=796886748550189
Ran key_averages() to make sure FunctionEvent code working as expected:
-- ------------ ------------
Name Self CPU % Self CPU CPU total % CPU total CPU time avg Self CUDA Self CUDA % CUDA total CUDA time avg # of Calls
ProfilerStep* 0.74% 3.976ms 64.40% 346.613ms 69.323ms 0.000us 0.00% 61.710ms 12.342ms 5
Optimizer.zero_grad#SGD.zero_grad 0.76% 4.109ms 0.76% 4.109ms 821.743us 0.000us 0.00% 0.000us 0.000us 5
## forward ## 6.89% 37.057ms 27.19% 146.320ms 29.264ms 0.000us 0.00% 58.708ms 11.742ms 5
aten::conv2d 0.22% 1.176ms 7.74% 41.658ms 157.199us 0.000us 0.00% 27.550ms 103.962us 265
aten::convolution 0.79% 4.273ms 7.52% 40.482ms 152.762us 0.000us 0.00% 27.550ms 103.962us 265
aten::_convolution 0.69% 3.688ms 6.73% 36.209ms 136.637us 0.000us 0.00% 27.550ms 103.962us 265
aten::cudnn_convolution 6.04% 32.520ms 6.04% 32.520ms 122.719us 27.550ms 8.44% 27.550ms 103.962us 265
aten::add_ 2.42% 13.045ms 2.42% 13.045ms 30.694us 12.700ms 3.89% 12.700ms 29.882us 425
aten::batch_norm 0.19% 1.027ms 8.12% 43.717ms 164.971us 0.000us 0.00% 16.744ms 63.185us 265
aten::_batch_norm_impl_index 0.31% 1.646ms 7.93% 42.691ms 161.096us 0.000us 0.00% 16.744ms 63.185us 265
------------------------------------------------------- ------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------
Differential Revision: D55925068
Pull Request resolved: https://github.com/pytorch/pytorch/pull/123650
Approved by: https://github.com/aaronenyeshi
Summary:
Kineto traces use microsecond level granularity because of chrome tracing defaults to that precision. Fix by adding preprocessor flag to TARGETS and BUCK files. Also remove any unnecessary ns to us conversions made in the profiler itself.
This diff contains profiler changes only. Libkineto changes found in D54964435.
Test Plan:
Check JSON and chrome tracing to make sure values are as expected. Tracing with flags enabled should have ns precision. Tracings without flags should be same as master.
Tracing with flags enabled: https://www.internalfb.com/intern/perfdoctor/trace_view?filepath=tree/traces/dynocli/devvm2185.cco0.facebook.com/rank-0.Mar_18_14_37_22.4155151.pt.trace.json.gz&bucket=gpu_traces
Tracing without flags enabled: https://www.internalfb.com/intern/perfdoctor/trace_view?filepath=tree/traces/dynocli/devvm2185.cco0.facebook.com/rank-0.Mar_18_14_39_15.4166047.pt.trace.json.gz&bucket=gpu_traces
Tracing on main: https://www.internalfb.com/intern/perfdoctor/trace_view?filepath=tree/traces/dynocli/devvm2185.cco0.facebook.com/rank-0.Mar_18_14_42_43.4177559.pt.trace.json.gz&bucket=gpu_traces
Ran key_averages() to make sure FunctionEvent code working as expected:
-- ------------ ------------
Name Self CPU % Self CPU CPU total % CPU total CPU time avg Self CUDA Self CUDA % CUDA total CUDA time avg # of Calls
ProfilerStep* 0.74% 3.976ms 64.40% 346.613ms 69.323ms 0.000us 0.00% 61.710ms 12.342ms 5
Optimizer.zero_grad#SGD.zero_grad 0.76% 4.109ms 0.76% 4.109ms 821.743us 0.000us 0.00% 0.000us 0.000us 5
## forward ## 6.89% 37.057ms 27.19% 146.320ms 29.264ms 0.000us 0.00% 58.708ms 11.742ms 5
aten::conv2d 0.22% 1.176ms 7.74% 41.658ms 157.199us 0.000us 0.00% 27.550ms 103.962us 265
aten::convolution 0.79% 4.273ms 7.52% 40.482ms 152.762us 0.000us 0.00% 27.550ms 103.962us 265
aten::_convolution 0.69% 3.688ms 6.73% 36.209ms 136.637us 0.000us 0.00% 27.550ms 103.962us 265
aten::cudnn_convolution 6.04% 32.520ms 6.04% 32.520ms 122.719us 27.550ms 8.44% 27.550ms 103.962us 265
aten::add_ 2.42% 13.045ms 2.42% 13.045ms 30.694us 12.700ms 3.89% 12.700ms 29.882us 425
aten::batch_norm 0.19% 1.027ms 8.12% 43.717ms 164.971us 0.000us 0.00% 16.744ms 63.185us 265
aten::_batch_norm_impl_index 0.31% 1.646ms 7.93% 42.691ms 161.096us 0.000us 0.00% 16.744ms 63.185us 265
------------------------------------------------------- ------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------ ------------
Differential Revision: D55087993
Pull Request resolved: https://github.com/pytorch/pytorch/pull/122425
Approved by: https://github.com/aaronenyeshi
Summary:
Move the profiler's Approximate Clock from libtorch to libc10. The main reason is to allow c10 features to get time.
The clock is using TSC when available for performance. CUDA Caching Allocator's implementation of memory snapshot will add the timestamps to memory events with this same clock in subsequent diff.
Test Plan: CI
Differential Revision: D50601935
Pulled By: aaronenyeshi
Pull Request resolved: https://github.com/pytorch/pytorch/pull/111972
Approved by: https://github.com/davidberard98
Summary:
## About Sync Events
For CUDA profiling mode, we can enable tracing CUDA synchronization events.
* This feature captures synchronization events in CUDA including 1) context/device sync, 2) stream sync, 3) CUDA event sync, 4) CUDA stream wait event (inter stream synchronization). Read more
* We add this flag using the profiler's experimental config option.
* This PR relies on 7b003638c6 change in pytorch/kineto
## Usage
Just set the `enable_cuda_sync_events` option in `_ExperimentalConfig`
```
from torch.autograd.profiler import profile, _ExperimentalConfig
with profile(use_kineto=True, use_cuda=True,
experimental_config=_ExperimentalConfig(enable_cuda_sync_events=True),
) as prof:
workload()
```
**Please wait for PyTorch github repo to point to 7b003638c6 or later commit in Kineto**
Test Plan:
## Unit Test
Added a unit test
buck2 test mode/dev-nosan caffe2/test:profiler --local-only -- test_profiler_cuda_sync_events
Tests finished: Pass 1. Fail 0. Fatal 0. Skip 0. Build failure 0
ttps://www.internalfb.com/intern/testinfra/testrun/281475298097379
Reviewed By: davidberard98
Differential Revision: D46244591
Pull Request resolved: https://github.com/pytorch/pytorch/pull/105187
Approved by: https://github.com/aaronenyeshi
1. `torch.autograd.profiler` interface parameters changed. (use `self.use_device` instead of `self.use_cuda` facilitates access by other devices and integrate it in subsequent pr)
2. Modify `ProfilerEventStub`(aka `std::shared_ptr<CUevent_st>`) to `ProfilerVoidEventStub`(aka `std::shared_ptr<void>`) so that `ProfilerStubs` can be inherited by any `{device}Methods`.
In addition, `cuda_event_start_` is renamed to `device_event_start_` , cuda and other devices can use this event pointer if needed.
4. custom device support using legacy profiling(add `ProfilerState::KINETO_PRIVATEUSE1_FALLBACK` option)
5. add `privateuse1Stubs` register
(parse results and test cases are added in subsequent pr)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/101554
Approved by: https://github.com/aaronenyeshi
Many ops take as inputs scalars or scalar lists which are important to understand the properties of the op. For example, convolution ops' behavior and output shapes often depend on padding and strides, which are provided as scalars of lists of scalars. This will record scalar lists when record_inputs=True.
Details:
During collection (and this was true before this PR as well), we serialize values and tensor metadata into an InputOutputEncoder. After collection occurs, we deserialize these values to attach the information to each of the events.
This PR does this:
- Adds support for serializing scalar lists during collection / serialization
- Adds an extra field called "Concrete Args"
- Splits up the deserialization process into two steps - one for generating "input shapes" and one for generating "concrete args". We split up input shapes and concrete args to avoid interrupting any previous workflows that relied on the specific data in the input shapes category; additionally, it's just a better description. Note that single scalars will remain in the "input shapes" category as they were already in that category in the past.
Differential Revision: [D45798431](https://our.internmc.facebook.com/intern/diff/D45798431)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/100593
Approved by: https://github.com/aaronenyeshi
Summary: This allows an internal use case to register a callback that can vary over time instead of being a static value over the lifetime of the program.
Test Plan: ran the test listed above ^^.
Differential Revision: D45805139
Pull Request resolved: https://github.com/pytorch/pytorch/pull/101292
Approved by: https://github.com/aaronenyeshi
Summary: We don't think the performance impact of recording concrete shapes is significant; but it's good to have a knob for turning it off quickly in case it has a large performance impact.
Test Plan:
Ran D45681838. It prints the state of that "concrete inputs" boolean. I ran it before and after canarying a change to `pytorch/kineto:pytorch_record_concrete_inputs`; before, it returns true; after, it returns false.
Note that D45681838 had to add `service` on the main function. That's because we need to `initFacebook` in order to use jks.
Differential Revision: D45680162
Pull Request resolved: https://github.com/pytorch/pytorch/pull/101043
Approved by: https://github.com/aaronenyeshi
