frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 12:54:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

Record the XPU and XCCL build settings in the compiled binary (#147161)

Browse Source 
Fixes #ISSUE_NUMBER

Currently the XPU and XCCL build settings are not recorded in the compiled binary and are not shown using the `torch.__config__.show()` which is a quick way to check if the binary has been built with such support.

Below is the output adding them (see end of last line):

```
Python 3.12.8 | packaged by conda-forge | (main, Dec  5 2024, 14:24:40) [GCC 13.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> print(torch.__config__.show())
PyTorch built with:
  - GCC 13.3
  - C++ Version: 201703
  - Intel(R) oneAPI Math Kernel Library Version 2025.1-Product Build 20250203 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v3.5.3 (Git Hash 66f0cb9eb66affd2da3bf5f8d897376f04aae6af)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - CPU capability usage: AVX512
XPU backend  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=RelWithDebInfo, COMMIT_SHA=43eb39d7c832b5560f7bfa8d29cc7919ac21c0ca, CXX_COMPILER=/home/pkourdis/compilers/gcc-13.3.0/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=1 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DUSE_KINETO -DLIBKINETO_NOCUPTI -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=OFF -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=range-loop-construct -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-unknown-pragmas -Wno-unused-parameter -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-dangling-reference -Wno-error=dangling-reference -Wno-error=redundant-move -DUSE_XPU -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, TORCH_VERSION=2.7.0, USE_CUDA=0, USE_CUDNN=OFF, USE_CUSPARSELT=OFF, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=1, USE_MPI=0, USE_NCCL=OFF, USE_NNPACK=0, USE_OPENMP=ON, USE_ROCM=0, USE_ROCM_KERNEL_ASSERT=OFF, USE_XCCL=1, USE_XPU=1,
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/147161
Approved by: https://github.com/guangyey, https://github.com/EikanWang, https://github.com/albanD

Co-authored-by: Yu, Guangye <106960996+guangyey@users.noreply.github.com>
This commit is contained in:
Panagiotis Kourdis
2025-05-20 09:21:39 +00:00
committed by PyTorch MergeBot
parent 1075bb37d3
commit 44f19c7179
2 changed files with 22 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
caffe2/core/macros.h.in
View File

@ -29,7 +29,7 @@
#endif
// Useful build settings that are recorded in the compiled binary
// torch.__build__.show()
// torch.__config__.show()
#define CAFFE2_BUILD_STRINGS { \
{"TORCH_VERSION", "${TORCH_VERSION}"}, \
{"CXX_COMPILER", "${CMAKE_CXX_COMPILER}"}, \
@ -65,4 +65,6 @@
{"USE_ITT", "${CAFFE2_USE_ITT}"}, \
{"USE_ROCM_KERNEL_ASSERT", "${USE_ROCM_KERNEL_ASSERT}"}, \
{"USE_CUSPARSELT", "${USE_CUSPARSELT}"}, \
{"USE_XPU", "${USE_XPU}"}, \
{"USE_XCCL", "${USE_XCCL}"}, \
}

19
test/test_xpu.py
View File

@ -1,5 +1,6 @@
# Owner(s): ["module: intel"]
import re
import subprocess
import sys
import tempfile
@ -732,6 +733,24 @@ class TestXPUAPISanity(TestCase):
if not torch.xpu._is_compiled():
self.assertEqual(len(torch.xpu.get_arch_list()), 0)
def test_torch_config_for_xpu(self):
config = torch.__config__.show()
value = re.search(r"USE_XPU=([^,]+)", config)
self.assertIsNotNone(value)
if torch.xpu._is_compiled():
self.assertTrue(value.group(1) in ["ON", "1"])
value = re.search(r"USE_XCCL=([^,]+)", config)
if torch.distributed.is_xccl_available():
self.assertTrue(value.group(1) in ["ON", "1"])
else:
self.assertTrue(value.group(1) in ["OFF", "0"])
else:
self.assertTrue(value.group(1) in ["OFF", "0"])
self.assertFalse(torch.distributed.is_xccl_available())
value = re.search(r"USE_XCCL=([^,]+)", config)
self.assertIsNotNone(value)
self.assertTrue(value.group(1) in ["OFF", "0"])
if __name__ == "__main__":
run_tests()