Threads inside the thread pools are not named, so they inherit the main process name or the name of the first thread. In our case if we set `pt_main_thread` as the thread name when a thread does `import torch`, this name will be inherited by all the threads in the created pools.
This PR names the threads in the pools I was able to find. There are other pools created, like OpenMP ones and we need to follow-up on those.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130270
Approved by: https://github.com/d4l3k, https://github.com/albanD
If `cpuinfo_initalize` returns false, call to subsequent cpuinfo functions may result in `abort()`
Also, `defaultNumThreads()` method is assumption if one method fails then try another, and finally return 1.
Alas there are no good way to test it on x86 platform, but on ARM one can replicate it by running `sudo chmod 750 /sys` and then `python3 -c "import torch;torch._C.profiler.gather_traceback(True, True, True)"`
<!--
copilot:poem
-->
### <samp>🤖 Generated by Copilot at 4d942e8</samp>
> _`cpuinfo` fails_
> _avoid undefined behavior_
> _check before you count_
Partially addresses https://github.com/pytorch/pytorch/issues/113568
Pull Request resolved: https://github.com/pytorch/pytorch/pull/113771
Approved by: https://github.com/atalman
The "import torch" crashes with following cpuinfo error on powerpc64.
==============================================================
>>> import torch
Error in cpuinfo: processor architecture is not supported in cpuinfo
Fatal error in cpuinfo: cpuinfo_get_processors_count called before cpuinfo is initialized
Aborted (core dumped)
==================================================================
The patch fixes this by excluding powerpc from using cpuinfo as it is not supported for ppc64.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/110708
Approved by: https://github.com/ezyang
This PR prefers "logical processor number" (the cpu cores shown in htop) returned by cpuinfo for determining c10 thread number. If that fails, it uses hardware_concurrency exactly.
The motivation is that in a x86 host with 64 cores and Hyper-Threading disabled, the current behavior uses 32 threads, resulting half of cores being idle.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/107010
Approved by: https://github.com/ezyang
Not only is this change usually shorter and more readable, it also can yield better performance. size() is not always a constant time operation (such as on LinkedLists), but empty() always is.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/93236
Approved by: https://github.com/malfet
Summary:
As GoogleTest `TEST` macro is non-compliant with it as well as `DEFINE_DISPATCH`
All changes but the ones to `.clang-tidy` are generated using following script:
```
for i in `find . -type f -iname "*.c*" -or -iname "*.h"|xargs grep cppcoreguidelines-avoid-non-const-global-variables|cut -f1 -d:|sort|uniq`; do sed -i "/\/\/ NOLINTNEXTLINE(cppcoreguidelines-avoid-non-const-global-variables)/d" $i; done
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/62008
Reviewed By: driazati, r-barnes
Differential Revision: D29838584
Pulled By: malfet
fbshipit-source-id: 1b2f8602c945bd4ce50a9bfdd204755556e31d13
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/58457
This variable had concurrent read/write access without any
synchronization. The issue was caught and reported by TSAN.
ghstack-source-id: 129311384
Test Plan:
1) Verify test locally.
2) waitforbuildbot.
Reviewed By: ezyang
Differential Revision: D28498116
fbshipit-source-id: 89af068467fed64c131d743504c0cecf3017d638
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/56830
Opt into formatting on GitHub and format everything. This is a trial run before turning on formatting for more and eventually all of the codebase.
Test Plan: CI
Reviewed By: zertosh
Differential Revision: D27979080
fbshipit-source-id: a80f0c48691c08ae8ca0af06377b87e6a2351151
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/37681
By passing by value, we can std::move, and avoid unnecessarily copying
args that are part of any std::function/lambda state (e.g. in the jit
interpreter, there is a std::vector<> stack passed in the
InterpreterContinuation)
This makes the api also consistent with e.g. folly and best practices.
Added a minor at::launch() benchmark to test/cpp/, the difference is
mostly noticeable when copying the std::function<> internal args is
non-trivial.
Benchmarks pre/post (min over ~5 runs)
NoData: 5.81 us -> 5.63 us (-3.2%)
WithData(0): 6.67 us -> 5.88 us (-11.8%)
WithData(4): 6.98 us -> 6.51 us (-6.7%)
WithData(256): 9.44 us -> 7.89 (-16.5%)
ghstack-source-id: 103322321
Test Plan:
- perf: buck run mode/opt caffe2/test/cpp/api:parallel_benchmark pre/post
- correctness buck test mode/dev-nosan caffe2/test/...
Reviewed By: dzhulgakov
Differential Revision: D21355148
fbshipit-source-id: 3567e730845106f1991091e4a892d093e00571c3
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/33938
Making sure we don't silently ignore exceptions from the tasks in the
thread pool
Test Plan: python setup.py clean && python setup.py develop install
Differential Revision: D20178603
Pulled By: ilia-cher
fbshipit-source-id: 34971032205a1a53fb7419ed84ebb986f9e959ad
Summary:
This improves locality and affinity by keeping work on the same
threads preferentially to starting work on new ones, and reduces
contention on the threadpool lock more generally.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/17808
Differential Revision: D14391282
Pulled By: resistor
fbshipit-source-id: 3aec81656a50460a725aa4187c61864295d4f46e
Summary:
1. Move ATen threadpool & open registration mechanism to C10
2. Move the `global_work_queue` to use this open registration mechanism, to allow users to substitute in their own
Pull Request resolved: https://github.com/pytorch/pytorch/pull/17788
Reviewed By: zdevito
Differential Revision: D14379707
Pulled By: jamesr66a
fbshipit-source-id: 949662d0024875abf09907d97db927f160c54d45
