On ROCm, using a non-vectorized index_put kernel provides ~2x perf improvement over the hipified CUDA kernel. None of the existing unit tests were exercising the large index case so a new unit test was added.
It was also noted that the scale value in the original kernel was hard-coded to 1.0 which would be a no-op, so it was removed from the simplified rocm kernel.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/138259
Approved by: https://github.com/xw285cornell, https://github.com/leitian, https://github.com/eqy
The `usort` config in `pyproject.toml` has no effect due to a typo. Fixing the typo make `usort` do more and generate the changes in the PR. Except `pyproject.toml`, all changes are generated by `lintrunner -a --take UFMT --all-files`.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/127126
Approved by: https://github.com/kit1980
The `usort` config in `pyproject.toml` has no effect due to a typo. Fixing the typo make `usort` do more and generate the changes in the PR. Except `pyproject.toml`, all changes are generated by `lintrunner -a --take UFMT --all-files`.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/127126
Approved by: https://github.com/kit1980
ghstack dependencies: #127122, #127123, #127124, #127125
This replaces a bunch of unnecessary lambdas with the operator package. This is semantically equivalent, but the operator package is faster, and arguably more readable. When the FURB rules are taken out of preview, I will enable it as a ruff check.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/116027
Approved by: https://github.com/malfet
Fixes incorrect range check when index is `std::numeric_limits<int64_t>::min()`, as result of unary minus operations for such values is undefined, but in practice is equal to self, see https://godbolt.org/z/Wxhh44ocr
Lower bound check was `size >= -index`, which was incorrect if `index` is `INT64_MIN`, with `-1 - index`, which for all int64_t values returns result that also fits into int64_t range. `- (index + 1)` is more readable and results in the identical optimized assembly, see https://godbolt.org/z/3vcnMYf9a , but its intermediate result for `INT64_MAX` is outside of `int64_t` range, which leads to a similar problems as with `int64_min` in original example.
Added regression test.
Fixes https://github.com/pytorch/pytorch/issues/115415
Pull Request resolved: https://github.com/pytorch/pytorch/pull/116062
Approved by: https://github.com/Skylion007, https://github.com/albanD
- Replace `for inst in instructions[0:targe.offset//2]: inst.starts_line = None`, with the one that that iterates over all instructions until `inst.offset == target.offset` condition is met, this way making it uniform across Python bytecode dialects (Python-3.11+ bytecode size is variable, while bytecode size is fixed for older Pythons)
- Speedup target_index search by replacing `[i for i in instructions if i.offset == offset][0]` with `next(i for i in instructions if i.offset == offset)`, which aborts the evaluation after condition met for the first time, according to:
```python
In [1]: lst=list(range(10000))
In [2]: %time [i for i in lst if i == 10]
CPU times: user 144 µs, sys: 23 µs, total: 167 µs
Wall time: 168 µs
Out[2]: [10]
In [3]: %time next(i for i in lst if i == 10)
CPU times: user 6 µs, sys: 0 ns, total: 6 µs
Wall time: 9.06 µs
Out[3]: 10
```
- Fix small typo
- use `is_py311_plus` variable rather than checking `sys.version_info`
<!--
copilot:poem
-->
### <samp>🤖 Generated by Copilot at 6cd7f27</samp>
> _We fix the typos in our code of doom_
> _We remove the warnings that obscure our vision_
> _We refactor the `generate` function for the dynamo_
> _We resume the execution with precision_
Fixes https://github.com/pytorch/pytorch/issues/103355
Pull Request resolved: https://github.com/pytorch/pytorch/pull/103525
Approved by: https://github.com/Skylion007, https://github.com/williamwen42
Fixes#99665
Let me explain the root cause using the unit test I added:
* This bug is triggered when:
* ```wrapped``` is a nested function.
* ```wrapped``` is in another module which is different from the main function ```fn```.
* There is a graph break inside of ```wrapped```.
* The root cause is when resuming nested function, actually we are using the outermost function(```fn``` in my example)'s global variables, but ```wrapped``` calls ```inner_func``` which is not part of ```fn```'s globals, so we have to set correct globals when nested function resume execution.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/100426
Approved by: https://github.com/jansel
Fixes#99665
Let me explain the root cause using the unit test I added:
* This bug is triggered when:
* ```wrapped``` is a nested function.
* ```wrapped``` is in another module which is different from the main function ```fn```.
* There is a graph break inside of ```wrapped```.
* The root cause is when resuming nested function, actually we are using the outermost function(```fn``` in my example)'s global variables, but ```wrapped``` calls ```inner_func``` which is not part of ```fn```'s globals, so we have to set correct globals when nested function resume execution.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/100426
Approved by: https://github.com/jansel
This prevents us from guarding on leading unbacked SymInts.
The previous attempt at https://github.com/pytorch/pytorch/pull/94521 I got the logic a bit wrong. My idea there was to avoid slicing when the values to be set have low enough dimensionality that they definitely aren't too long. To do this, I need to compute the difference between the data to be set, and the post-slice space for the values. But I incorrectly compared against the *pre-slice* space in the original PR. Another version of this PR which is wrong is to compare against variableIndices.size(); but remember that in advanced indexing with tensors/lists, each of the individual indices specify what coordinates to read out of each dimension! A third incorrect attempt tested `variableIndices[0].dim()`, which is only correct if you don't broadcast one of the later variable indices, and if there are enough variableIndices to cover all dims. This is all quite complicated, so I went for a simpler solution of checking if the leading dim had a hint before testing if it is not equal to one.
BTW, there is no test for this one stripping behavior. There is now a test for this, based off the real code that caused the problem.
Signed-off-by: Edward Z. Yang <ezyangmeta.com>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/95141
Approved by: https://github.com/ngimel
Rerun all disabled test to gather their latest result so that we can close disabled tickets automatically. When running under this mode (RERUN_DISABLED_TESTS=true), only disabled tests are run while the rest are skipped `<skipped message="Test is enabled but --rerun-disabled-tests verification mode is set, so only disabled tests are run" type="skip"/>`
The logic is roughly as follows, the test runs multiple times (n=50)
* If the disabled test passes, and it's flaky, do nothing because it's still flaky. In the test report, we'll see the test passes with the following skipped message:
```
<testcase classname="TestMultiprocessing" file="test_multiprocessing.py" line="357" name="test_fs" time="0.000" timestamp="0001-01-01T00:00:00">
<skipped message="{"flaky": True, "num_red": 4, "num_green": 0, "max_num_retries": 3, "rerun_disabled_test": true}" type="skip"/>
</testcase>
```
* If the disabled test passes every single time, and it is not flaky anymore, mark it so that it can be closed later. We will see the test runs and passes, i.e.
```
<testcase classname="TestCommonCUDA" name="test_out_warning_linalg_lu_factor_cuda" time="0.170" file="test_ops.py" />
```
* If the disabled test fails after all retries, this is also expected. So only report this but don't fail the job (because we don't care about red signals here), we'll see the test is skipped (without the `flaky` field), i.e.
```
<testcase classname="TestMultiprocessing" file="test_multiprocessing.py" line="357" name="test_fs" time="0.000" timestamp="0001-01-01T00:00:00">
<skipped message="{"num_red": 4, "num_green": 0, "max_num_retries": 3, "rerun_disabled_test": true}" type="skip"/>
</testcase>
```
This runs at the same schedule as `mem_leak_check` (daily). The change to update test stats, and (potentially) grouping on HUD will come in separated PRs.
### Testing
* pull https://github.com/pytorch/pytorch/actions/runs/3447434434
* trunk https://github.com/pytorch/pytorch/actions/runs/3447434928
Pull Request resolved: https://github.com/pytorch/pytorch/pull/88646
Approved by: https://github.com/clee2000
Currently index_select/index_add decompositions decompose to `index` or `index_put` ops. The problem with this is that `index_select` and `index_add` accept int32 indices while `index` doesn't. That leads to error in meta func for those decompositions. This PR adds non-performant support for int32 indices to `index` operations, to allow decompositions go through.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/86309
Approved by: https://github.com/lezcano
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/71708
In Python 3.2, a number of asserts were deprecated.
In Python 3.11, these asserts are deleted completely. The files in this change still use the deprecated asserts.
Switch over to the supported syntax for 3.2 onwards.
Test Plan: Tested on the internal test suite runner.
Reviewed By: ajtulloch
Differential Revision: D33503694
fbshipit-source-id: a150f296033260acf8365d77b837ce0679f57361
(cherry picked from commit abf60ed97409265222915d8265aaabedd625fd93)
Summary:
Description:
- Follow up PR to https://github.com/pytorch/pytorch/issues/66790 to fix the tests on functorch, https://github.com/pytorch/functorch/issues/195
In functorch, a null tensor is added to the list of indices for the batch dimension in C++, but I can not find an equivalent of that in python without using `torch.jit.script`. If any other better solutions could be suggested, I'd be happy to replace the current way of testing.
cc ngimel zou3519
Pull Request resolved: https://github.com/pytorch/pytorch/pull/67189
Reviewed By: suo
Differential Revision: D31966686
Pulled By: ngimel
fbshipit-source-id: a14b9e5d77d9f43cd728d474e2976d84a87a6ff4
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/67032
This PR adds meta backend support to the `range`, `arange`, `linspace`, and `logspace` operators.
Note that the original PR (#66630) was reverted due to two failing unit tests in the Bionic CI. This revision includes a fix for those tests; otherwise its content is identical to the previous PR.
Original commit changeset: 2f9d8d1acbb0
ghstack-source-id: 142487306
Test Plan: Extended the existing tensor creation tests to assert meta backend support.
Reviewed By: zhaojuanmao
Differential Revision: D31834403
fbshipit-source-id: a489858a2a8a38a03234b14408e14d2b208a8d34
Summary:
https://github.com/pytorch/pytorch/issues/57515
Based on ngimel 's branch, with a few tweaks to determine when to copy value tensors to device memory/additional tests.
bc-breaking note: Previously, if in `x[index]=value` `value` was a 0-d tensor with device different from `x`'s device, it resulted in a RuntimeError. Now this case is handled by copying `value` to the correct device.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/61612
Reviewed By: mrshenli
Differential Revision: D29753491
Pulled By: ngimel
fbshipit-source-id: 3fba14f4c2b9b136b50af020f9c1eda88f7373b0
Summary:
Let index/index_put implementation in aten take care of moving the indices to the correct device, don't make python wrapper do that.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/59059
Reviewed By: mruberry
Differential Revision: D28750562
Pulled By: ngimel
fbshipit-source-id: 2f2b5f875733898f1c0b30b544c89808f91e4a6f
Summary:
Reference: https://github.com/pytorch/pytorch/issues/38349
Wrapper around the existing `torch.gather` with broadcasting logic.
TODO:
* [x] Add Doc entry (see if phrasing can be improved)
* [x] Add OpInfo
* [x] Add test against numpy
* [x] Handle broadcasting behaviour and when dim is not given.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/52833
Reviewed By: malfet
Differential Revision: D27319038
Pulled By: mruberry
fbshipit-source-id: 00f307825f92c679d96e264997aa5509172f5ed1
